ENGIEERING
MATHEMATICS
VOLUMEI
Dr. Rajesh Pandey
MSc., Ph.D.
Assistant Professor/Reader
Department of Mathematics
Sherwood College of Engineering,
Research and Technology Lucknow,
Faizabad Road, Barabanki (U.P.)
Lucknow
Published by
wordpress
Khushnuma Complex Basement
7, Meerabai Marg (Behind Jawahar Bhawan)
Lucknow 226 001 V.P. (INDIA)
Tel.:915222209542,2209543,2209544,2209545
Fax: 05224045308
EMail: wordpress@hotmail.com
First Edition 2010
ISBN 9789380257037
Publisher
All Rights Reserved
No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior written permission of the author.
Composed & Designed at:
Panacea Computers
3rd Floor, Agrawal Sabha Bhawan
Subhash Mohal, Sadar Cantt. Lucknow226 002
Tel.:05222483312,9335927082,9452295008
Email: prasgupt@rediffmail.com
Printed at:
Salasar Imaging Systems
C7/5, Lawrence Road Industrial Area
Delhi 110035
Tel.:01127185653,9810064311
Basic Results and Concepts
I. GENERAL INFORMATION
1. Greek Letters Used
a alpha e theta
~ beta ~ phi
y gamma \jf psi
(5 delta
S xi
E epsilon TJ eta
i iota l; zeta
A lambda
2. Some Notations
E belot1.gs to uunion
n intersection => implies
<=> implies and implied
by
3. Unit Prefixes Used
Multiples and Prefixes
Submultiples
10
3
kilo
10
2
hecto
10 deca
10
1
deci*
10
2
centi*
10
3
milli
10
6
micro
K kappa "C tau
/l mu X chi
vnu (0 omega
7t pi r cap. gamma
P rho Ll caE delta
cr sigma L cap. sigma
~ doesnot belong to
I such that
Symbols
k
h
da
d
c
m
J.l
* The prefixes 'dedI and 'centi' are only used with the metre, e.g., Centimeter is a
recognized unit of length but Centigram is not a recognized unit of mass.
4. Useful Data
e = 2.7183 lie = 0.3679
7t = 3.1416 1/7t = 0.3183
J2 = 1.4142 J3 = 1.732
loge2 = 0.6931
loge10 = 2.3026
1 rad. = 5717'45"
viii
loge 3 = 1.0986
logH)e = 0.4343
10 = 0.0174 rad.
iii
5 S t . ,ys ems 0 fU 't nl s
Quantily F.P.s. System e.G.S. System M.K.S. System
Length foot (ft) centimetre (cm) metre(m)
Mass pound (lb) gram (gm) kilogram (kg)
Time second (sec) second (sec) second (sec)
Force lb. wt. dyne newton (nt)
6. Conversion Factors
1 ft. = 30.48 cm = 0.3048 m 1m = 100 cm = 3.2804 ft.
1 ft2 = 0.0929 m
2
1 acre = 4840 yd
2
= 4046.77 m
2
lft3 = 0.0283 m
3
1 m
3
= 35.32 ft3
1 ml sec = 3.2804 ftl sec. 1 mile Ih = 1.609 km/h.
II. ALGEBRA
1. Quadratic Equation: ax
2
+ bx + C = 0 has roots
b + !(b
2
 4ac)  b  ~ ( b 2  4ac)
a =" , p = '
2a 2a
b c
a + p =  , ap = .
a a
Roots are equal if b
2
 4ac = 0
Roots are real and distinct if b
2
 4ac > 0
Roots are imaginary if b
2
 4ac < 0
2. Progressions
(i) Numbers a, a + d, a + 2d ...... are said to be in Arithmetic Progression (A.P.)
f  n
Its nth term Tn = a + n  1 d and sum Sn =  (2a + n  1 d)
2
(ii) Numbers a, ar, ar
2
, ...... are said to be in Geometric Progression (G.P.)
1 a(l  rn) a
Its nth term T = ar
n
 and sum S = S = (r < 1)
n n 1r' '" 1r
(iii) Numbers l/a, 1/(a + d), 1/(a + 2d), .... are said to be in Harmonic Progression
(H.P.) (i.e., a sequence is said to be in H.P. if its reciprocals are in A.P. Its nth term
Tn =1/(a+n1d).)
(iv) If a and b be two numbers then their
Arithmetic mean = ! (a + b), Geometric mean = jiili; Harmonic mean = 2ab/(a
2
+ b)
(v) Natural numbers.are 1,2,3 ... ,n.
Ln = n(n + 1) Ln2 = n(n + 1) (2n + 1) Ln3 = {n(n
2
+ 1)}2
2 ' 6'
ix
iv
(vi) Stirling's approximation. When n is large n!  .J21tn . nn e
n
.
3. Permutations and Combinations
n In! np
nPr = .. nC = = _r
(n  r)!' r r ! (n  r) ! r !
n =n n =l=n
C
n
_
r
Cr I c" en
4. Binomial Theorem
(i) When n is a positive integer
(1 + x)n = 1 + nCt X + nC2 x
2
+ nC3 x
3
+ ....... + nC
n
xn.
(ii) When n is a negative integer or a fraction
(1+xt =1 +nx+ n(n 1)x
2
+ n(n1)(n  2)x
3
+ ..... 00.
1.2 1.2.3
5. Indices
(i) am . an = a
m
+
n
(ii) (am)n = a
mn
(iii) a
n
= l/a
n
(iv) n Fa (i.e., nth root of a) = a 1/n .
6. Logarithms
(i) Natural logarithm log x has base e and is inverse of ex.
Common logarithm lOglOX = M log x where M = lOglOe = 0.4343.
(ii) loga 1= 0; logaO =  oo(a > 1); loga a = 1.
(iii) log (mn) = log m + logn; log (min) = log m log n; log (mn) = n log m.
III. GEOMETRY
1. Coordinates of a point: Cartesian (x ,y) and polar (r , 8).
Then x = r cos 8, Y = r sin 8
or r= ~ ( x 2 + y2), 8 = tan
1
( ~ ) 
y
p
8
o
x x
x
v
Distance between two points
(XlI yd and (x2/Y2) = +(Y2Yl)2=]
Points of division of the line joining (XlI Yl) and (X21 Y2) in the ration ml : m2 is
(
ffilX2 + ffi 2Xl I ffi l Y2 + ffi2Yl )
ffi
l
+ ffi2 ffi
l
+ ffi2
In a triangle having vertices (XlI Yl), (X2, Y2) and (X31 Y3)
1 Xl Yl 1
(i) area =  x
2
Y2 1.
2 x3 Y3 1
(ii) Centroid (point of intersection of medians) is
(
Xl + X2 + X3 Yl + Y2 + Y3 )
3 I 3
(iii) Incentre (point of intersection of the internal bisectors of the angles) is
[
aXl + bX2 + cX3 I aYl + bY2 + CY3 J
a+b+c a+b+c
where a, b, c are the lengths of the sides of the triangle.
(iv) Circumcentre is the point of intersection of the right bisectors of the sides of
the triangle.
(v) Orthocentre is the point of intersection of the perpendiculars drawn from the
vertices to the opposite sides of the triangle.
2. Straight Line
(i) Slope of the line joining the points (XlI Yl) and (X21 Y2) = Y2  Yl
51 f th 1
b 0 . a . coeff ,
of
X
ope 0 e me ax + Y + c = IS  I.e. I  
b eoeff/of Y
(ii) Equation of a line:
X2  Xl
(a) having slope m and cutting an intercept c on yaxis is Y = mx + c.
(b) cutting intercepts a and b from the axes is + .r = 1.
a b
(c) passing through (XlI Yl) and having slope m is Y  Yl = m(x  Xl)
(d) Passing through (XlI Y2) and making an La with the X  axis is
X Xl _ Y  Yl _
 r
cos a sin a
(e) through the point of intersection of the lines alx + bly + Cl = 0 and a2X + h2y +
C2 = 0 is alX + blY + Cl + k (a2x + b2Y + C2) = 0
(iii) Angle between two lines having slopes ml and m2 is tanl ffi
l
 ffi2
1 ffi
l
ffi
2
xi
vi
Two lines are parallel if ml = m2
Two lines are perpendicular if mlm2 = 1
Any line parallel to the line ax + by + c = 0 is ax + by + k = 0
Any line perpendicular to ax + by + c = 0 is bx  ay + k = 0
(iv) Length of the perpendicular from (Xl, Yl)of the line ax + by + c = O. is
aXl + bYl + c
+ b
2
) .
y
o
3. Circle
(i) Equation of the circle having centre (h, k) and radius r is
(x  h)2 + (y  k)2 = r2
x
(ii) Equation X2 + y2 + 2gx + 2fy + c = 0 represents a circle having centre (g, f)
and radius = + f2  c).
(iii) Equation of the tangent at the point (Xl, Yl) to the circle x2 + y2 = a
2
is XXI + yyl
= a
2
.
(iv) Condition for the line y = mx + c to touch the circle
X2 + y2 = a
2
is c = a + m
2
).
(v) Length of the tangent from the point (Xl, Yl) to the circle
x2 + y2 + 2gx + 2fy + C = 0 is  + 2gxl + 2fyl + c).
4. Parabola
(i) Standard equation of the parabola is y2 = 4ax.
Its parametric equations are X = at
2
, y = 2at.
Latus  rectum LL' = 4a, Focus is S (a,O)
Directrix ZM is X + a = O.
xii
vii
y
M b+;:?I
o
II
ctl
+
X
Z A
(ii) Focal distance of any point P (XII YI ) on the parabola
y2 = 4ax is SP = Xl + a
(iii) Equation of the tangent at (Xl' YI) to the parabola
y2 = 4ax is YYI = 2a (x + Xl)
(iv) Condition for the line Y = mx + c to touch the parabola
~ = 4ax is c = aim.
x
(v) Equation of the normal to the parabola y2 = 4ax in terms of its slope m is
y = mx  2am  am
3
.
5. Ellipse
(i) Standard equation of the ellipse is
x2 y2
+=1
a
2
b
2
.
M
L
y
M'
B P (XI y)
~  ~   +         ~       ~     ~   ~ x
Z C Z'
L'
B'
xiii
viii
Its parametric equations are
x = a cos 8, Y = b sin 8.
Eccentricity e = b
2
/ a
2
) .
Latus  rectum LSL' = 2b
2
/ a.
Foci S ( ae, 0) and S' (ae, 0)
Directrices ZM (x =  a/e) and Z'M' (x = a/e.)
(ii) Sum of the focal distances of any point on the ellipse is equal to the major axis
i.e.,
SP + S'P = 2a.
(iii) Equation of the tangent at the point (Xl' Yl) to the ellipse
x
2
y2 xx yy
+=lis
l
+_1 =1
a
2
b
2
a
2
b
2
.
(iv) Condition for the line y = mx + c to touch the ellipse
.i. =1 is c = + b
2
).
a
2
b
2
6. Hyperbola
(i) Standard equation of the hyperbola is
x2 y2
  =1
a
2
b
2
.
Its parametric equations are
x = a sec S, y = b tan 8.
Eccentricity e = /a=2),
Latus  rectum LSL' = 2b
2
/ a.
y
M'
Z' C
Directrices ZM (x = a/e) and Z'M' (x =  a/e).
M
Z
(ii) Equation of the tangent at the point (Xl' Yl) to the hyperbola
xiv
x
ix
2 2
~ _ L = 1 is XXI _ YY I = 1.
a
2
b
2
a
2
b
2
(iii) Condition for the line y = mx + c to touch the hyperbola
~  L = 1 is c = J(a
2
m
2
_ b
2
)
a
2
b
2
. x
2
y2 X Y X Y
(lV) Asymptotes of the hyperbola    = 1 are  +  = 0 and    o.
a
2
b
2
a b a b
(v) Equation of the rectangular hyperbola with asymptotes as axes is xy = c
2
. Its
parametric equations are x = ct, Y = c/ t.
7. Nature of the a Conic
The equation ax
2
+ 2hxy + by2 + 2gx + 2fy + c = 0 represents
a h g
(i) a pair of lines, if h b f (= L'l) = 0
g f c
(ii) a circle, if a = b, h = 0, L'l :t: 0
(iii) a parabola, if ab  h
2
= O,e ~ :t: 0
(iv) an ellipse, if ab  h
2
> 0, L\:t: 0
(v) a hyperbola, if abh
2
> 0, L\:t: 0
and a rectangular hyperbola if in addition, a + b = O.
8. Volumes and Surface Areas
Solid Volume Curved Surface
Area
Cube (side a) a
3
4a
2
Cuboid (length 1, Ibh 2 (l + b)h
breadth b, height
h)
Sphere (radius r)
4 ::I

 ref'
3
Cylinder (base 1tr2h 2nrh
radius r, height
h)
Cone 1
nr
2
h
nrl
3
where slant height I is given by I = J(r
2
+ h
2
).
xv
Total Surface
Area
6a
2
2 (Ib + bh + hI)
41tr2
2nr (r + h)
1tr (r + 1)
x
IV. TRIGONOMETRY
1.
9
0
= 0 0 30
sin 9 0 1/2
cos 9 1
13

2
tan 9 0
1/13
45 60 90 180
1/ J2 13
1 0

2
1/ J2
1/2 0 1
1
13
00 0
2. Any tratio of (n. 90
0
8) = same ratio of 8, when n is even.
= co  ratio of 8, when n is odd.
270 360
1 0
0 1
00 0
The sign + or  is to be decided from the quadrant in which n. 90
0
8 lies.
1
e.g., sin 570
0
= sin (6 x 90
0
+ 300) = sin 30
0
=  2'
tan 315
0
= tan (3 x 90
0
+ 45
0
) =  cot 45
0
=  1.
3. sin (A B) = sin A cos B cos A sin B
cos (A B) = cos A cos B sin A sin B
sin 2A = cos
2
A cos A = 2 tan A/(J + tan
2
A)
2 2 2 2 1  tan
2
A
cos 2A = cos A  sm A = 1  2 sm A = 2 cos AI = .
tanAtanB 2 tan A
4. tan (A B) = ; tan 2A = 2'
1 + tan A tan B 1  tan A
5. sin A cos B = ..!.. [sin (A + B) + sin (A  B)]
2
cos A sin B = ..!.. [sin (A + B)  sin (A  B)]
2
1
coa A cos B =  [cos (A + B) + cos (A  B)]
2
sin A sin B = ..!.. [cos (A  B)  cos (A + B)].
2
6
. C . D 2' C+D C D
.sm +sm = smcos
2 2
C+D C D
sinC  sinD=2cos sin
2 2
C+D CD
cos C + cos D = 2 cos cos 
2 2
. C+D . CD
cosC  cos D=  2sm sm
2 2
7. a sin x + b cos x = r sin (x + 8)
a cos x + b sin x = r cos (x  8)
xvi
1 + tan
2
A
xi
where a = r cos, a, b = r sina so that r = J(a
2
+ b
2
), a tan
1
(b/a)
8. In any ~ A B C :
(i) a/sin A = b/sin B = c/sin C (sin formula)
b
2
+ c
2
_ a
2
(ii) cos A = . (cosine formula)
2bc
(iii) a = b cos C + C cos B (Projection formula)
(iv) Area of ~ A B C = .!bc sin A = Js(s  a) (s  b) (s  c) where s = .!(a + b + c).
2 2
9. Series
2 3
(i) Exponential Series: eX = 1 + ~ + ~ + ~ + ...... 00
I! 2! 3!
(ii) sin x, cos x, sin hx, cos hx series
x
3
x
5
sinx=x   +   ...... 00,
3! 5!
x
2
X4
cos X = 1   +   ...... 00
2! 4!
x
3
x
5
sin h x = x +  +  + ...... 00,
3! 5!
(iii) Log series
x
2
X4
cos h x = 1 +  +  + ..... 00
2! 4!
x
2
x
3
log (1 + x) = x  +   ..... 00,
2 3
[
x2 x
3
)
log (1  x) =  x + "2 + "3 + .... 00
(iv) Gregory series
x
3
x
5
1 1 + X x
3
x
5
tan
1
x=x  +   ..... 00,tanh
1
x= log =x+ + + .... 00.
3 5 2 Ix 3 5
10. (i) Complex number: z = x + iy = r (cos a + i sin a) = re
i6
(ii) Euler1s theorem: cos a + i sin a = e
i9
(iii) Demoivre1s theorem: (cos a + isin a)n = cos na + i sin n a.
11. (i) Hyperbolic functions: sin h x = eX  e
x
;cos h x = eX + e
x
;
2 2
sin h x cos h xII
tan h x = . cot h x = . sec h x = . cosec h x == 
cosh x ' sin h x ' cos h x ' sin hx
(ii) Relations between hyperbolic and trigonometric functions:
sin ix = i sin h x i cos h x = cos h x i tan ix = i tan h x.
(iii) Inverse hyperbolic functions;
sinh
1
x==log[x+j;;2;1];cosh
1
x=log[x+Jx
2
1];tanh
1
x==.:!..log l+x.
2 1  x
xvii
xii
V.CALCULUS
1. Standard limits:
n n
(i) Lt .x  a = nanI,
x+a x a
n any rational number
(iii) Lt (1 + x)l/x = e
x+o
aX  1
(v) Lt = logea.
x+ 0 X
2. Differentiation
. d dv du
(1)  (uv) = u  + v 
dx dx dx
du du dy .
 = .  (cham Rule)
dx dy dx
(ii) = eX
dx
d
 (loge x) = l/x
dx
(
"') d (. )
111  SIn x = cos x
dx
d
 (tan x) = sec
2
x
dx
d
 (sec x) = sec x tan x
dx
(iv) (sin
1
x) = 1
dx (1 _ x
2
)
(tan
1
x) = _1_
dx 1 + x
2
(sec
1
x) = 1
dx  1)
(v) (sin h x) = cos h x
dx
(ii) Lt sin x = 1
x+ 0 x
(iv) Lt xlix = 1
x+ <TJ
= v du / dx  u dv / dx
dx V v
2
(ax + b)" = n (ax + b)n1. a
dx
d
t<a
X
) = aX logea
d 1
 (Iogax) =
dx x log a
(cos x) =  sin x
dx
d
 (cot x) =  cosec
2
x
dx
d
 (cosec x) =  cosec x cot x.
dx
d 1 1
 (cos x) = t====
dx _ x
2
)
d 1 1
 (cot x)=
dx 1 + x
2
d 1
 (cosec
1
x) = ;===
dx 1) .
(cos h x) = sin h x
dx
(tan h x) = sech
2
x (cot h x) = h
2
x.
dx dx
(vi) Dn (ax + b)m = m (m 1) (m  2) ...... (m  n + 1) (ax + b)m  n . an
Dn log(ax + b) = ( 1) n1 (n  1) ! an/(ax + b)n
Dn (e
ffiX
) = mne Dn (a
mx
) = mn (loga)n. affiX
xiii
Dn [Sin (ax : b)] = (a2 + b2)n/2 eax [Sin(bX + c + n tan
1
b / a) ].
cos{bx c) cos(bx+c+ntan1b/a)
(vii) Leibnitz theorem: (UV)n
= Un + nC1Un1Vl+ nC2Un_2V2 + ..... + nCrUn_rV
r
+ ..... + nCnVn.
3. Integration
n + 1 1
(i) fx
n
dx = ; + 1 (n:!; 1) f; dx = loge
x
fex dx = eX fax dx = aX jlogea
(ii) fSin x dx =  cos x fcos x dx = sin x
ftan x dx =  log cos x fcot x dx = log sin x
fsec x dx = log(sec x + tan x) = log tan + %)
fcosec x dx = log(cosec x  cot x) = log tan (%)
fsec
2
x dx = tanx
(
... ) f dx 1 t 1 X
111 =  an
a
2
+ x
2
a a
J
dx 1 I a + x
a2 _ x2 = 2a og a  x
J
dx 1 I x  a
, =  og
x
2
_ a
2
2a a + x
_ a2) dx = x  a
2
) + a
2
coshJ = _ a2) _ log ___ a_2_)
2 2 a 2 2 a
(v) Je
ax
sin bx dx = 2 e
ax
2 (a sin bx  b cos bx)
a + b
J
eax
e
ax
cos bx dx = 2 2 (a cos bx + bsin bx)
a + b
xix
xiv
(vi) fsin h x dx = cos h x
ftan h x dx = log cos h x
f cos h x dx = sin h x
fcot h x dx = log sin h x
fsec h
2
x dx = tan h x fcosech
2
x dx =  cot h x.
r/2 ("/2
(vii).b sinn x dx =.b cos
n
x dx
(n  1) (n  3) (n  5) .... (1t l'f' )
= x , on yIn IS even
n(n  2) (n  4) 2
[
/2 . m n d (rn  1) (rn  3) .... x (n  1) (n  3) ....
sm x cos x x = ~    '   ~   '     '       '  '     '  
(rn + n) (rn + n  2) (rn + n  4) ....
x ( ~ , only if both rn and n are even J
(viii) r f (x) dx = r f(a  x) dx
r f(x) dx = 2 r f(x) dx, if f (x) is an even function.
= 0, if f(x) is an odd function.
L
2a
f(x) dx = 2 r f(x) dx, if f(2a  x) = f(x)
= 0, if f(2a  x) =  f(x).
VI. VECTIORS
1. (i) IfR=xi+
y1
+zkthenr= IRI =J(x
2
+y2 +z2)
(ii) PQ = position vector of Qposition vector of P.
2. If A = a1 I + a2J + a3K, B = b1I + b2J + b, then
(i) Scalar product: A . B = abcos8 = a
1
b
1
+ a2b2+ a3b3
(ii) Vector product: A x B = ab sin 8 N = Area of the parallelogram having A + B
as sides
I J K
= a
1
a
2
a,
b
1
O
2
0
3
(iii) BA 1. IF A . B = and A is parallel to B if Ax B =
a
1
a? a,
3. (i) Scalar triple product [A B C] = b
1
b; 0
3
= Volume of parallelopiped
(1 (2 (,
(ii) If [A B C] = 0, then A, B, C are coplanar
(iii) Vector triple product A x (B x C) = (A . C) B  (A . B) C
(A x B) x C = (C . A) B  (C . B) A
xx
xv
. Of Df Df
4.(1) gradf=Vf= 1+ J+K
Ox ay GZ
. aft af
2
dlV F = VF =  +  + '
dx ay ?z
. . .
i . k
curl F = V x F = 1 where F = ftl + f2J + bK
Ox av ()z
f
t
t; f3
(ii) If div F = 0, then F is called a solenoidal vector
(iii) If curl F = 0 then F is called an irrotational vector
5. Velocity = dR/ dt; Acceleration = d
2
R/ dt2;
Tangent vector = dR/dt; Normal vector =
6. Green's theo<em , l' + ",dy) " I Ie (z  : 1 dxdy
Stoke's theorem: JF. dR = JcurI F. N ds
c s
Gauss divergence theorem: JF . N ds = JdiV F dv
s v
7 C d' t t . oor tna e sys ems
Polar coordinates Cylindrical
(r,8) coordinates (p, <p , z)
Coordinate x = r cos 8 x = p cos
transformations y = r sin e y = p sin
z=z
Jacobian a(x , v) 8(x, y, z)
 = r =p
D( r, 8) D(p, <p, z)
(Arc  length)2 (dS)2 = (drF + r2 (ds)2= (d p)2 + p2
(d8F + (dZ)2
dx dy = rd8 dr
Volume element dV = pdp dq, dz
xxi
Spherical polar
coordinates (r, 8,
x = r sin8 cos
y = r sin 8 sin
Z = r cos 8
8(x, y, z) 2 8
= r SIn
8(r, 8,<p)
(ds)2 = (dr)2 + r
2
(d8)2 + (r sin 8)2
(dq, )2
dV = r2 sin 8 dr d8
dq,
xvi
Contents
Chapters
Basic Results and Concetps
Unit I : Differential Calculus I
Chapter 1 : Successive Differentiation and Leibnitz's Theorem
Chapter 2: Partial Differentiation
Chapter 3: Curve Tracing
Chapter 4 : Expansion of Function
Unit II: Differential Calculus  II
Chapter 5: Jacobian
Chapter 6 : Approximation of Errors
Chapter 7 :Extrema of Functions of Several Variables
Chapter 8 : Lagranges Method of Undetermind Multipliers
Unit III: Matrices
Chapter 9: Matrices
Unit IV : Multiple Integers
Chapter 10 : Multiple Integres
Chapter 11 : Beta and Gamma Functions
Unit V: Vector Calculus
Chapter 12: Vector Differential Calculus
Chapter 13 : Vector Integral Calculus
Page No.
3
19
51
83
95
105
115
127
139
191
223
251
305
"This page is Intentionally Left Blank"
UNITl
Differential CalculusI
"This page is Intentionally Left Blank"
Chapter 1
Succesive Differentiation and
Leibnitz's Theorm
Successive Differentiation
Definition and Notation : If y be a function of x, its differential coefficient
dy / dx will be in general a function of x which can be differentiated. The
differential coefficient of dy / dx is called the second differential coefficient of y.
Similarly, the differential coefficient of the second differential coefficient is called
the third differential coefficient, and so on. The successive differential coefficients
of yare denoted by
dy d
2
y d
3
y
dx ' dx2 ' dx3 ................ .
the nth differential coefficient of y being dny
dxn
Alternative methods of writing the nth differential coefficient are
(
d )n dny
dx y, Dny, Yn, dxn ,y(n)
The Process to find the differential coefficient of a function again and again is
called successive Differentiation.
Thus, if y = f(x), the successive differential coefficients of f(x) are
dy d
2
y d
3
y dny
dx' dx2 ' dx3 , .................... dx"
These are also denoted by :
(i) yl, y2, y3 ............ yn
(
.. ) I II lit n
11 y, Y ,y ............ y
(iii) Dy, D2y, D3y ............ Dny
(iv) f'(x), f"(x), f"'(x), ............ .fn(x)
nth Derivatives of some standard functions :
(1) nth derivative of (ax+b)m :
Let y = (ax+b)m
Differentiating it w.r.t. x in succession, we get
yl = m(ax+b)ml.a
Y2 = m(ml)(ax+b)m
2
a
2
y3 = m(ml) (m2) (ax+b)m3 a
3
3
A Textbook of Engineering Mathematics Volume  I
Similarly, we can write
yn = m(m1) (m2) (ax+b)rnn an
Dn (ax+b)rn = m(m1) (m2) ......... (mn+1) (ax +b)mn. an
If m is a positive integer then R.H.S of the above can be written as
I!!! (ax +b)rnn. an
I(mn)
Hence Dn(ax+b)rn = I!!! .a
n
(ax +b)mn
I(mn)
Deductions:
(a) If m = n, then Dn(ax+b)n = l!! .a
n
In particular, Dn xn = l!!
Dn2 xn = l!! x2
~
Dn3 xn = ~ x
3
etc.
(b) If m = p, where p is a positive integer, then
I(p+n 1)
Dn(ax+b)p = (l)n l(p1) an (ax+b)pn
(
1)n
1
n
(c) D
n
(ax+br
1
= I.!! an
(ax + b)"+1
(d) If m< n, then Dn (ax+b)rn =0
2. nth Derivative of e
ax
:
Lety = e
ax
Differentiating w.r.t. x in succession we get
yl = a ea
x
, y2 = a
2
ea
x
, y3 = a
3
e
ax
,
Similarly we can write yn = an e
ax
:. Dn e
ax
= an e
ax
3. nth derivative of a
RlX
:
Lety = a
RlX
Differentiation w.r.t x in succession, we get
Yl = ma
RlX
log a
Y2 = m
2
arnx (log a)2
Similarly, we can write yn = mn a
RlX
(log a)n
:. Dn arnx = mn a
RlX
(log a)n
In particular,
Dn ax = ax (log a)n
4. nth Derivative of log (ax+b)
Lety = log (ax+b)
4
Succesive Differentiation and Leibnitz's Theorm
a
yl =  = a(ax+b)l
ax+b
Y2 = (1) (ax+b)2 a
2
y3 = (1) (2) (ax+b)2 a
3
Similarly we can write
yn = (1) (2) ......... (n+1) (ax+b)n an
= (1)n
1
1.2.3 ................ (n1) (ax+b)n an
= (_1)nl I(n 1). (ax+b)n. an
=
(ax + b)"
(_1)"1 a"l(n 1)
:. On log (ax+b) = ===
(ax + b)"
5. nth Derivative of sin (ax+b):
Let y = sin (ax+b)
:. yl = a cos (ax+b)
Yl = a sin (ax+b+ 2: )
2
n
Y2 = a
2
cos (ax+b+  )
2
Y2 = a
2
sin (ax+b+ 2n)
2
y3 = a
3
sin (ax+b+ 3n )
2
Similarly, we can write
yn = an sin (ax+b+ nn)
2
:. On sin (ax+b) = an sin(ax+b+ nn)
2
In particular,
O
n ( nn)
sln X = sln x+
2
6. nth Derivative of cos (ax+b) :
Lety = cos (ax+b)
:. Yl = a sin (ax+b)
n
= a cos (ax+b+)
2
Y2 = a
2
sin (ax+b+ 2: )
2
5
A Textbook of Engineering Mathematics Volume  I
n
= a
2
cos (ax+b+2.  )
2
y3 = a
3
sin (ax+b+ 2n )
2
3n
= a
3
cos (ax+b+ )
2
Similarly, we can write
nn
yn = an cos (ax+b+)
2
n nn
:. 0 cos (ax+b) = an cos (ax+b+)
In particular,
n nn
:.0 cos x = cos (x+)
2
7. nth Derivative of e
ax
sin (bx+c)
Let y = e
ax
sin (bx +c)
2
Yl = a e
ax
sin (bx+c)+ b e
ax
cos (bx+c)
put a = r cos<l> and b = r sin<l>
:. r2 = a
2
+b
2
and <I> = tan
1
(bja)
Yl = r e
ax
[cos<l> sin (bx+c) +sin<l> cos (bx+c)]
Similarly, we can obtain
Y2 = r2 e
ax
sin (bx+c+ 2<1
y3 = r3 e
ax
sin (bx+c+ 3<1
Continuingthis process n times, we get
yn = rn e
ax
sin (bx+c+n<l
:. On {eax sin (bx+c)} = rn. e
ax
sin(bx+c+n<l
where r = (a
2
+b
2
)1/2 and <I> = tan
1
( ~ )
8. nth Derivative of e
ax
cos (bx+c) :
Proceeding exactly as above, we get
On (e
ax
cos(bx+c)}= rn. e
ax
cos (bx+c+n<l
where r and <I> have the same meaning as above.
Example 1: if y = 1 2 ' find yn
15x+6x
Solution:
1
y = 1 5x+6x
2
6
(U.P.T.U.2005)
Succesive Differentiation and Leibnitz's Theorm
1
=
(2x 1)(3x 1)
2 3
=
2x1 3x1
:. yn = 2Dn (_1_) 3Dn (_1 )
2x1 3x1
= 2[(1)n l!}(2)n ]_3[(1)" l!}(3)"]
(2x _1)"+1 (3x 1r+
'
= 1 n n Ans.
[
(2)"+1 (3),,+1]
( ) l!} (2x _l)n+l (3x 1r+
1
2
Example 2: if y = ~ , find nth derivative of y.
(x1) (x+2)
Solution: To split y into partial fractions
Let x1 = z, then
1 1+2z+z2
y=2"'
z 3+z
1 (1 5z 4 Z2 )
= Z2 "3+9+9" 3+z
1 5 4
=++
3z
2
9z 9(3 + z)
1 5 4
= + +
3(x _1)2 9(x 1) 9(x + 2)
Hence
(_l)n I(n + 1) 5(1)" l!} 4(1)"l!}
Y
= + + Answer.
n 3(x _1)n+2 9(x _1)"+1 9(x + 2)n+l
Example 3: Find the n'" derivative of t a n ~ l ((1 ::'))
Solution : we have
tanI ( 2x ) = 2 tanIx
(1 x
2
)
Hence we have to find yn if y = 2 tanIx
Now, y = 2 tanIx
7
(U.P.T.U.2002)
(U.P.T.U. 2001)
A Textbook of Engineering Mathematics Volume  I
Yt = l:x
2
=  X:iJ
 x:J
Hence,
 1 (l)ntl( 1)[ 1 1 1
Y n  :  n  ( .)n  ( . )n
I XI X+I
Suppose X = r case and 1 = r sine so that
r2=x2+ 1 and e = tan
l
(.;)
Thus,
Yn = (I)";;nl!!..::1{(coseisinern (cosS+isinern}
(_I)n
1
In 1
=> Y n = . n .2i sin ne
I.r
=> Yn = 2(I)"11(n 1).sinnS.sin
n
S
Where S = tan
l
(.;) Answer
xI
Example 4 : If Y = x log , show that
x+l
n
2
1 [ X  n x + n 1
Yn =(1) (n2) (xI)"  (x+l)"
Solution :
Y = x[ log(xl) log (x+l)] (i)
Differentiating Y w.r.t x, we get
YI = x(_I ___ 1_)+log (xI) log (x+l)
xI x+l
(U.P.T.U 2003)
1 1
= ++ log (xI) log (x+l) (ii)
xI x+l
Again, Differentiating (ii), (nl) times w.r.t. x we gets
(I)"21(n  2) (_I)n
2
1(n  2)
Yn = (xl)n + (x+l)" + (xl)"l  (x+l)"l
=(_I)"21(n_2)[(I)(nI)+(I)(nI)+ xI _ x+1 ]
(xI)" (x+l)n (xI)" (x+l)"
8
Succesive Di(ferentiation and Leibnitz's Theorm
 ] Answer
Leibnitz's Theorem :
Statement : If u and v are functions of x then
Dn (uv)=(Dn u).v + nCl Dnl u. Dv + nC2 Dn2 u. D
2
v + ............ + nCr Dnr u. Drv
+ .......... +u.Dnu
Proof: We shall prove this theorem by mathematical induction method. By
differentiation, we have
D(uv) = (Du). v + u. (Dv)
=(Du). v + (ICl U ) (Dv)
Thus the theorem is true for n = 1.
Let the theorem be true for n = m i.e.
Dm (uv) = (Dmu). v + mel Dml u. Dv + mC2 Dm2 u. D
2
v + ......... +m
Cr
Dmru. Drv
+ .................. + uDmv
Differentiating the above once again, we get
Dm+l(uv) = (Dm+lu). v+Dmu. DV+
m
Cl{D
m
u. Dv+Dm
l
u. D
2
v}+me2{Dm
l
u.D2v+
Dm
2
u. D3v}+ ........... +mer{Dmr+lu. Drv+Dmru. Dr+lv}+ .......... + ......... +{Du.
D
m
v+uDm+lv}
Rearranging the terms, we get
Dm+l(uv) = (Dm+lu)v+(1+mCl) D
m
u.Dv+(mCl+
m
C2) Dm
l
u. D
2
v+ ........... + (mCr
+mer+l) Dmru. Dr+lv+ ............ +u. Dm+lv.
Now using mCr + mCr+l = m+1C
r
+
l
, we have
Dm+l(uv)= (Dm+lu).v +m+1CIDmu.Dv+ ......... +m+1Cr+1Dmru.Dr+l v+ ......... +
u.Dm+lv
Thus we have seen that if the theorem is true for n = m, it is also true for n = m+l.
Therefore by principle of induction, the theorem is true for every positive integral
value ofn.
Example 5: If y = cos(m sin
I
x) prove that
(1_X2) yn+2  (2n+1) Xyn+l + (mL n2)Yn =0
Solution: y = cos (m sin
I
x)
:. Yl =  sin (m sinIx). m Jb
1x
2
or (1_X2) y
1
2 = m
2
sin
2
(m sin
I
x)
= m
2
[1  cos
2
(m sin
1
x)]
= m
2
 m
2
cos
2
(m sin
I
x)
:. (1  X2) y12 = m
2
 m2y2
or (1_X2) y12 + m2y2_ m
2
=0
Differentiating again, we have (1x
2
). 2Y1Y2+Y1
2
(2x)+m2 . 2YYI =0
or (1  X2) Y2 Xyl + m2y = 0
9
A Textbook of Engineering Mathematics Volume  I
Differentiating n times by Leibnitz's theorem, we get
yn+2 (1x2)+ nCl yn+l (2x) + nC2 yn(2)  {Yn+l. x+ n
C1
yn.1} +m
2
yn =0
or (1x2)Yn+2  2nx yn+l  n(n1)Yn  xy n+l  nyn + m2Yn =0
or (1x2)Yn+2  (2n+1) xy n+l + (m2n2) = 0 Hence proved.
I If
asinlx h th .
Examp e 6 : y = e ,s ow at (1x
2
) yn+2  (2n+1)xy n+l  (n2+a
2
) yn =0
Solution: Given y = easmlx (i)
Differentiating,
asin1x 1
yl= e a
_ ay

Squaring and crossmultiplying, we get
Y12(1x2) = a
2
y2
Differentiating both sides with respect to x, we get
Y12(2x) + 2YlY2(1X2) =a
2
.2YYl
(ii)
or (1_X2)Y2XYI  a
2
y =0 (iii)
Differentiating each term of this equation n times with respect to x, we get
Dn[(1x2)Y2]  Dn(x Yl)  a
2
Dn(y) =0.
n(n 1)
or [(1x2)Yn+2 + nyn+l(2x)+ yn (2)]  [XYn+l + nyn.l]  a
2
Yn =0
1,2
or (1x2) yn+2  (2x+1) Xyn+l  (n2 + a
2
) yn = 0
Hence proved.
Example 7: If y = (x
2
1)", Prove that
(x2.1)Yn+2 +2XYn+l  n(n+l)Yn =0
Solution: Given y = (x2.1)n
Differentiating, yl = n(x
2
1)nl. 2x
or Yl(x2.1) = n(x21)nl. 2x
= ny. 2x
=2nxy
Differentiating again both sides with respect to x, we get
Yl(2x) + yz(xL1) = 2n(xYl+Y)
or (x
2
1) Y2 + 2x(1n)Yl  2ny = 0
Differentiating each term of this equation n times with respect to x (by Leibnitz's
theorem), we get
Dn[(xL1)Y2]+Dn[2x(1  n)y]  2nDny =0
n(n 1)
or [(xL1)Yn+2 +nyn+l (2x)+ yn(2)]+ 2(1  n)[Xyn+l+nYn]  2nyn=0
or (X2 1)Yn+2 + 2XYn+l  n(n+1)Yn =0 Hence proved
10
Succesive Di[ferentiation and Leibnitz's Theorm
Example 8: If cos
1
(t) Prove that
X2Yn+2 + (2n+1)xYn+1 + 2n2Yn =0
Solution: Given cos
1
(y jb) = log(xj n)n
or cos
1
(y jb) = n log(xjn)
or y = b cost n log(xjn)}
Differentiating,
y1 = b sin {n log (xjn)} n. _( 1 ) . .!.
x/n n
or Xyl =  b n sin{n log (x/ n)}
Again differentiating both sides with respect to x, we get
XY2 + y1 = bn cos {n } n. _( 1 ) . .!.
n x/n n
or X2Y2 + Xy1 =  n
2
b cos { n log (x/n)}
= n2y, from (i)
or x
2
y2 + Xyl + n2y = 0
Differentiating each term of this equation n terms, with respect to x, we get
Dn(x2Y2)+Dn (XYl) +n
2
Dn (y) =0
n(n 1)
or yn+2 x
2
+ nyn+l (2x) + yn (2) + (Yn+l x + nYn) + n
2
yn =0
1.2
or x
2
yn+2 + (2n+1) XYn+l + 2n
2
yn =0 Hence proved
Example 9: If x= cos h y] Prove that
(X2 1)Y2 + Xyl  m2y =0
and (X2 1) yn+2 + (2n+1) xyn+l + (n2  m2) yn =0
Solution: Here x = cos h )log y]
1
or cos hl x =  log y
m
or m cos hl x = log y
ory =
emcos h
I
"
:.y! m[
=my
or (X2 1) Y1
2
= m
2
y2
Again differentiating (ii) we get
11
(i)
(ii)
A Textbook of Engineering Mathematics Volume  I
(x2..1) 2Yl y2 + (2X)Yl2 = m
2
. 2YYl
or (X2 1) y2 +XYl  m2y = 0 (iiH
Differentiating (iii) n times by Leibnitz's theorem, we get
n(n 1)
[(X2..1)Yn+2 + n (2X)Yn+2+ (2) yn] + [xy n+1 + n(1)YnJ  m
2
yn =0
or (x21) yn+2 + (2n+1) Xyn+l + (n2  m2) yn =0 Hence Proved
Example 10: If yl/m +y l/m = 2x Prove that (X2 1) yn+2 + (2n +1)XYn+l+(n2  m2) =0
Solution: Given y l/m + _1_ = 2x
yl/nt
or y2/m  2xyl/m + 1 =0
4)
or y 11m = "''
2
or yl/m = X 1)
ory=
when y =[ 1) r ,we have
2x 1
,,(x
2
1)
or Yl 1) = my
or Y1
2
(X2 1) = m2y2
wheny =[ 1) r, we have
2x 1
1)
=
m[ x 1) r
N=1
or yl 1) =  my
(ii)
or Y12(xL1) = m2y2 (iii)
which is the same result as (ii).
Hence for both the values of y given by (i) we get Y1
2
(X2 1) = m2y2
Differentiating both sides of this with respect to x, we get
Y1
2
(2x) + 2Y1Y2 (X2 1) = m2.2YY1
or Y2 (x
2
1) + Xyl  m2y =0
12
(i)
Succesive Ditferentiation and Leibnitz's Theorm
Differentiating each term of this equation n times (by Leibnitz's theorem) with
respect to x, we get
Dn{Y2(x2_1)}+Dn(xYl)  m
2
Dn(y) =0
n(n 1)
or {(X2 1) yn+2 +nyn+l (2x) + yn (2)} + {XYn+l + nYn.1}  m
2
yn =0
1.2
or (x
2
1) yn+2 + (2n +1) Xyn+l + (n2  m2) yn =0 Hence Proved
Example 11 : If y = (X2 1)n Prove that (X2 1)Yn+2 + 2XYn+l  n(n+1)Yn =0
d
n
Hence IfP
n
= (x21)n, show that
dxn
d {( 2) dP
n
}
 1x  +n(n+1)Pn=0
dx dx
Solution: Given y = (x
2
_1)n
Differentiating
yl = n(xL1)nl. 2x
or yl (X2 1) = n (X2 1)n 2x
or Yl(X
2
1) = ny.2x
= 2nxy
Differentiating again both sides with respect to x, we get
Yl(2x) +Y2 (x
2
1) = 2n [XYl +y]
or (x
2
1) Y2 + 2x(1 n) yl  2ny =0
Differentiating each term of this equation n times with respect to x (by Leibnitz's
theorem), we get
Dn{(x2 1)Y2} +Dn{2x(1 n)y} 2nDn y) =0
or [(xL1)Yn+2 + nyn+l (2x) + n(n 1) yn(2)] + 2(1  n){XYn+l + nyn} 2nyn =0
1.2
or (X2 1)Yn+2 + 2XYn+l  n (n+1)yn =0 (i)
which Proves the first part.
Again P
n
= ~ ( X 2 1)"
dxn
d
n
= dxn (y)
=yn
, d {( 2) dP
n
} _ d {( 2) }
,, 1x   1x 'Yn+1
dx dx dx
',' P
n
= yn
= (1X2)Yn+2  2XYn+l
= _[(X2 1)Yn+2 + 2x Yn+l]
=  n(n+1)Yn from (i)
=  n (n+1)P
n
',' P
n
= yn
13
A Textbook of Engineering Mathematics Volume  I
or  (1_x2)_n + n(n+1) Pn=O. Henceproved
d { dP }
dx dx
Problem 12 : If y = sin
1
x , find (yn)o
Solution: Given y = sin
1
x
Differentiating, y, = J 1
(1 x
2
)
or Y12(1 x
2
) =1
Differentiating both sides again w.r.t x we get
2YlY2 (1  X2) + Y1
2
(2x) =0
or Y2(1 X2)  XYI =0 ......... (iii)
(V.P.T.V.2009)
(i)
(ii)
Differentiating both sides of (iii) n times with respect to x by Leibnitz's theorem,
we get
n(n1)
[yn+2(1x
2
)+n.yn+l(2x)+ yn (2)]  [XYn+l +n.1. yn] =0
or (1X2)Yn+2  (2n+1)XYn+l  n2Yn =0 (iv)
putting x = 0 in (i), (ii), (iii) and (iv) we get
(y)o = sin
1
0 =0; (Yl)O =1; (Y2)0 =0
and (Yn+2)0 = n
2
(yn)0 (v)
If n is even, putting n = 2, 4, 6 ............ , we get
(Y4)0 =22 (Y2)0 = 0 :. (Y2)0 =0
(Y6)0 = 4
2
(Y4)0 = 0, :. (Y4)0 =0
In this way we can show that for all even values of n, (Yn)O =0 Answer
If n is odd, putting n=1,3,5,7 ............ (n2) in (v) we get
(Y3)0 = (Yl)O = 1; (ys)o = 3
2
(Y3)0 = 3
2
.1, (Y7)0 = 5
2
.(ys)o = 5
2
.3
2
.1 etc
(yn)o = (n  2)2 (yn 2)0
= (n  2)2. (n  4)2 (Yn 4)0
= (n  2)(n  4)2 .......... 5
2
,3
2
, 12. Answer.
Example 13 : If y = [log{x+ J(1 + x
2
) }P, show that
(Yn+2)0 =  n
2
(yn)O, hence find (Yn)O.
Solution: Let y = [log x+ J(1 + x
2
) F (i)
:.y, =2[IOg{X+
J
(1+X')}]{ 1 )[1+
2 1+x
or y, = [J(l !x,JIOg{X+J(l +x')}]
14
Succesive Differentiation and Leibnitz's Theorm
or y ~ (1 +x
2
) = 4 [log {x+ ~ ( 1 + x
2
) })2
or yt
2
(1+x
2
)= 4y (ii)
Again differentiating,
yt
2
(2x) + (1 +X2) 2YtY2 = 4yt
or y2 (1 +x
2
) + xyt 2 =0 (iii)
Differentiating each term of this equation n times, by Leibnitz's theorem, with
respect to x, we have
n(n 1)
{Yn+2 (1 +X2) + nyn+t (2x)+ Yn.2} + (yn+t(X) + nyn(l) } =0
1.2
or (1+x2)Yn+2 + (2n+1) x)'n+t +n2Yn = 0 ........... (iv)
putting x =0 in (iv) we have
(1 +0) (Yn+2)0 + n
2
(yn)O =0
or (yn+2)0 =  n2(yn)o ................. (v)
Putting n = n2, n4, n6 ......... .in (v), we get
(yn)o =  (n  2)2 (y n 2)0;
(yn 2)0 =  (n  4)2 (yn 4)0;
(Yn4)0 =  (n  6)2(Yn_6)0;
where we have (on multiplying side wise)
(yn)O = {(n  2)2} {  (n  4)2} { (n  6)2} {(yn 6)0}
:. If n is odd,
(yn)O = { (n  2)2}{  (n  4)2} {  (n  6)2} .................. { 3
2
} { 12}(Yt)o
Now from (i) putting x =0, we get
(y)o = [log (0+1)]2 = (log 1)2 =0
:. From (ii) putting x =0, we get
( Y l ) ~ = 4(y)0 =0 or (yt)o =0
Hence when n is odd (Yn)O =0 Answer
And if n is even, as before
(yn)o = { l(n  2)2}{ (n  4)2} { (n  6)2} ................ {4
2
} {_22}(Y2)0
Also from (iii) putting x =0, we get
(Y2)0 (1+0) +0  2 =0
or (Y2)0 =0
:. where n is even, we have
(yn)o = {  (n2)2} {  (n4)2} {  (n  6)2} ............. (2)2}2
= (_1)(n2)/2 (n2)2 (n4)2 (n6) ...... .4
2
.22.2 Answer
Example 14 : If y = [ x + ~ ( 1 + x
2
) r ' find (yn)O
Solution: Given y = [x + ~ ( 1 + x
2
) r ........... (i)
15
A Textbook ofEngineerins Mathematics Volume  I
:. y, =m[ x +)(1 +X2) ]",1 [1 + ~ 1
[
) 2 J
m
_
1
[)(1+X
2
)+X]
=m x+ (1 +x ) ~
(1 + x
2
)
=
my
~ ( 1 + x
2
)
Again differentiating
yi (2x) + (1 +X2) 2YlY2 = m
2
. 2YYl
(1
+ 2) + 2 0 ( ... )
or Y2 x Xyl  m y  ................... 111
Differentiating this equation n time by Leibnitz's theorem, we get
n(n 1)
[yn+2 (1+x2) + nyn+l (2x)+ yn (2)] + [XYn+l + nyn]  m
2
Yn =0
1.2
or (1 +X2)Yn+2 +(2n+ 1) Xyn+l + (n2  m2)Yn =0
pu tting x =0 we get
(yn+2)O = (m2  n2) (yn)O ........................... (iv)
If n is odd, putting n = 1,3, 5, 7 ............. (n2), we get
When n = 1, (Y3)O = (m2 12) (Yl)O
And from (i) and (ii), putting x =0, we get
(y)o = 1; ( Y l ) ~ = m
2
( y ) ~ = m
2
or (Yl)O = m
:. (Y3) 0 = (m212) m
when n = 3, from (iv) we get (ys)o = (m2 3
2
)(Y3)O
or (ys)o = (m2  3
2
) (m2 1) m
when n =5, from (iv) we get (Y7)O = (m2 52) (ys)o
or (Y7)O = (m2 52) (m2 3
2
) (m2 12)m.
Proceeding in this manner, when n = n  2 from (iv) we get
(yn)O = {m2  (n2)2} (Yn2)O
= {mL (n  2)2} {m2  (n  4)2} {m2  (n  6)2} ................ (m2 3
2
) (m2 12) m Answer.
If n is even, putting n = 2,4,6,8, ............... (n2) is (iv), we get
when n =2, (Y4)O = (m2  22) (Y2)O
And from (iii) putting x =0, we get
(Y2)O +0+m
2
(y)o = 0 or (Y2)O = m
2
(y)o = m
2
:.(Y4)O = (m
2
 22)m2
When n =4, from (iv) we get (Y6)O = (m2  42) (Y4)O
or (Y6)O = (m2  42) (m2  22) m
2
16
Succesive Differentiation and Leibnitz's Theorm
when n =6, from (iv), we get (ys)o = (m2  6
2
) (Y6)0
or (ys)o = (m
2
6
2
) (m2  42) (m2 22)m2
Proceeding in this manner, when n = n2 from (iv) we have
(yn)O = {m
2
 (n2)2} (Yn 2)0
={m2  (n2)2} {m2  (n4)2} ................ (m
2
 42) (m2 22)m2.
Answer
Exercise
1. If e
mcos
' x , prove that
(1x2) yn+2 = x (2n+1) yn+!  (n2 + m2) yn = 0
2. If Y = sin (a sino! x) prove that
(1x2) yn+2 (2n +1) xyn+!  (n2a
2
) yn = 0
3. If Y = a cos (log x) + b sin (log x) show that
X2Yn+2 + (2n +1) xyn+l (n2+1) Yn = 0
4. If Y = sin (m sino! x) prove that
(1x2) yn+2  (2n+1) XYn+! + (m2..n2)y n = 0
5. If Y = emcos 'x, show that
(1x2)Yn+2  (2n+1) xy n+1  (m2+n2) Yn = 0
and hence evaluate (Yn)O
U.P.T.U (CO) 2007
(U.P.T.U 2004)
(U.P.T.U 2003.05)
m1t
Ans: (yn)O = {(n2)2 +m2} {(n4)2+m2} ........... {(J2+m2) me}, for odd values of n
2
m1t
(yn)O = {{n2)2 +m
2
} {(n4)2+m
2
} .............. {m2e
T
} for even values of n.
17
"This page is Intentionally Left Blank"
Chapter 2
Partial Differentiation
Partial differential coefficients: The Partial differential coefficient of f(x,y) with
respect to x is the ordinary differential coefficient of f(x,y) when y is regarded as a
constant. It is written as
of
 or Of/ Ox or Dxf
ox
Thus af = lim f(x + h, y)  f(x, y)
ox h ..... O h
Again, the partial differential coefficient 0 f/ Oy of f(x,y) with respect to y is the
ordinary differential coefficient of f(x,y) when x is regarded as a constant.
Th
of l' f(x.y+k)f(x,y)
us  = 1m   '   "   '   ~  =  ' : "
oy k ..... O k
Similarly, if f is a function of the n variables Xl, X2, ......... Xn, the partial differential
coefficient of f with respect to Xl is the ordinary differential coefficient of f when
all the variables except Xl are regarded as constants and is written as Of/ Oxl.
of of .
 and  are also denoted by fx and fy respectively.
ox oy
The partial differential coefficients of fx and fy are f
xx
, fxy, f
yx
, fyy
02f 02f 02f 02f .
or 2' , 0'0 2' respectively.
ox oxoy oy X y
. 02f O(Of) 02f O(Of)
It should be speCIally noted that 0  means  0 and means   .
yox oy X oxoy ax oy
The student will be able to convince himself that in all ordinary cases
02f 02f
=
oyox oxoy
Example 1 : If u = log (X3 + y3 +Z3  3xyz) show that
(
0 0 0 )2 9
ox + ay + oz u =  (x + y + X)2
Solution: The given relation is
u = log(x3 + y3+ x
3
 3xyz)
19
(V.P.T.V. 2004, B.P.S.C. 2007)
(V.P.P.C.S.2oo3)
A Textbook of Engineering Mathematics Volume  I
Differentiate it w.r.t. x partially, we get
au 3x
2
3yz
ax = x
3
+ y3 + Z3  3xyz
'.. au 3y2 3xz
slmtlarly ~ , = 3 3 3
vy x' +Y' +z 3xyz
and au = 3z
2
 3xyz
az x
3
+ y3 + Z3  3xyz
au au au 3(x
2
+y2 +Z2 yzzxxy)
.  +  +  = '==::'''
.. ax Oy az x
3
+y3 +Z3 3xyz
_ 3(X2 +y2 +Z2 yzxzxy)
 (x + y + z) (X2 + y2 + Z2  yZ  zx  xy)
3
=
20
(u.p.p.es. 1998, B.P.S.C 2005)
Partial Differentiation
Solution: Given u = exYZ
au
. = xyexyz
.. az
a
2
u a a
_ = _( xye
XYz
) = x_ye
xyz
dyaz dy ay
= x[y xz eXYZ + exyz]
= exYZ (X2yz + x)
a
3
u a
Hence =  [exyz (X2yZ+X)]
axayaz ax
=eXYZ (2xyz+1) + yz exYZ (X2yZ+X)
= eXYZ [2xyz +1 + x2y2z2 + xyz]
= eXYZ (1 +3xyz + X
2
y
2Z2) Hence Proved.
x
2
y2 Z2
Example 3 : If 2 + 2 + 2 = 1, prove that I
a +u b +u c +u
u ~ + u ~ + u; = 2 (xu x +yUy + zUz)
2 2 2
Solution: Given ++ { + :: = 1 ................. (i)
a +u b +u c +u
where u is a function of x/y and z,
Differentiating (i) partially with respect to x, we get
2 2 au b2 2 au 2 2 au
(a +u).2xx  ( +u).Oy  (c +u).Oz 
____ ;;".:axc:,. + ax + ax = 0
~ 2 + u t ( ~ + u t ~ 2 + u t
2x [X2 y2 Z2 1 au
or a2 +u  (a
2
+ut + (b
2
+ut + (c
2
+ut ax =0
au 2x/(a
2
+ u)
or ax = [ x2 / (a2 + u t + y2 / (b
2
+ U)2 + Z2 / (c
2
+ U)2 J
_ 2xja
2
+u
 L[X
2
/(a
2
+utJ
.. au 2y/(b
2
+u)
Slmtlarly ;.},T = [ 2 J
vJ L x
2
/(a
2
+u)
au 2z / (c
2
+ u)
and  = =''=
az L[ x
2
/(a
2
+ utJ
21
(V.P.T.V.2003)
A Textbook of Engineering Mathematics Volume  I
:.(aU)2 +(aU)2 +(aU)2 =4[x
2
/(a
2
+ut +y2 /(b
2
+ut +:2 /(c
2
+utJ
, ax ay az [L{X2 /(a
2
+ut}]
2 2 2 4 ( .. )
or u. +U
y
+U
z
= L[{x
2
/(a
2
+ut}] ..................... 11
Also xUx + yu
y
+ zU
z
=
1 [2X2 2y2 2Z2 1
= L[ x
2
/(a
2
+utJ (a
2
+u) + (b
2
+u) + (c
2
+u)
= /(:' + u)'] [1] .................. (ili)
From (i), (ii) (iii) and we have
+ + u; = 2(xux +yuy + zUz) Hence Proved.
Example 4 : If u = f(r) and x = r cosS, y = r sinS i.e. r2 = X2+y2, Prove that
a
2
u a
2
u fll() 1 f'( )
+ r + r
ax
2
ay2 r
Solution: Given u = f(r) ................ (i)
Differentiating (i) partially w.r.t. x, we get
au = f'(r). or
ax ax
x
= f'(r).  .,' r2 = x
2
+y2
r
or
=2x
ax
or x
ax r
Similarly or = r
ay r
Differentiating above once again, we get
22
(V.P.T.V. 2000, 2005)
Partial Differentiation
a
2
u _ a [xf'(r)]
ax
2
 ax r
r[f'(r).l + xf"(r)(ar / ax)]  xf'(r)(ar / ax)
r2
a
2
u 1 x
2
or ax
2
=;r [rf' (r) + x
2
f"(r)  ;: f'(r)] (ii)
a
2
u 1 2
Similarly, 2 = 2" [rf'(r) + y2 fll(r)  L f'(r)] (iii)
dy r r
Adding (ii) and (iii), we get
2rf'(r)+(x
2
+y2)f"(r) f'(r)
a 2 a
2
1 [ (X2 + y2) 1
ax
2
ay2 r2 r
1
=  [2r f'(r) + r2 fll(r)  r f'(r)]
r2 .
1
=  f'(r) + fll(r), Hence proved.
r
Example 5 : If XX yy ZZ = c, show that at x = y =z,
a
2
z ( )1
= xlogex
axay
(V.P.P.C.S. 1994; P.T.V. 1999)
Solution: Given XX yy ZZ = c, where z is a function of x and y
Taking logarithms, x log x + Y log Y + z log z = log c (i)
Differentiating (i) partially with respect to x, we get
+ (log X)l] + + (log Z)l] = 0
az (1 +logx)
or  = 
ax (l+logz)
Similarly from (i) we have
az (l+logy)
ay =(l+logz)
:. = :x(
a [ (1+10
gy
)] F ("')
=  rom 111
ax 1 +logz
(ii)
(iii)
23
A Textbook Mathematics Volume  I
or =(l+logy)':x[(l+logz(]
= (1 + logy).[(l + logzt..!.. az]
z ax
or a
2
z = (1 + log y) 2 .{_(1 + log x)), using (ii)
axay z(l+logz) l+logz
a
2
z (1 +logx)2
At x = y = z, we have =  3
axOy x(1+1ogx)
Substituting x for y and z
. a
2
z 1
1 e =
.. axay x(l +Iogx)
1
:.log e = 1
= x(loge'+logx)
1
=
x log (ex)
=  {x log (ex)}l Hence Proved.
Example 6 : If u = sin
1
(f ) + tan
1
, show that
au au
x+y=o
ax ay
S I
. . W h ' 1 (x) + 1 (r) (.)
o uhon . eave u  sm y tan x' ............ 1
Differentiating (i) partially w.r.t. x and y, we get.
au 1 1 1 (y)
Ox = rn
oy
+ 0 ;1
24
(V.P.T.V.2006)
Partial Differentiation
x x
= +::7"
y _ X2 X
2
+ y2
au x xy ( ... )
or x ay =  _ x
2
+ x2 + y2 ................. 111
on adding (ii) and (iii), we have
au au
x+y=O Answer.
ax ay
Euler's Theorem on Homogeneous Functions:
Statement: If f(x,y) is a homogeneous function of x and y of degree n, then
af af
x+y=nf
ax ay
(U.P.T.V. 2006; P.T.V. 2004)
Proof: Since f(x,y) is a homogeneous function of degree n, it can be expressed in
the form
f(x,y) = xn F(y Ix) ... ................ (i)
:. af (r)(y)
ax ax x x
2
or x =n xn F  yx
n

1
F{ ................. (ii)
Again from (i), we have
af a
ay = ay (xnF(y/x)}
1
= xn P(y/x). 
x
af _ nl F'( I ) ( ... )
or y ay  yx y x .............. 111
Adding (ii) and (iii), we get
af af
x+ y = nxn F(y/x)
ax ay
= nf using(i) Hence Proved.
Note. In general if f (Xl, X2 ............... Xn) be a homogeneous function of degree n,
then by Euler's theorem, we have
25
A Textbook of Engineering Mathematics Volume  I
af af af
Xl aX
I
+ X, aX
2
+ .............. + Xn aX
n
= nf
(
x2 +y2)
Example 7 : If u = log , Prove that
x+y
au au
x+y=l
ax ay
Solution: We are given that
u = log (X2 + y2 )
x+y
x2 + y2
:. ell = = f(say)
x+y
Clearly f is a homogeneous function in x and y of degree 21 i.e. 1
:. By Euler's theorem for f, we should have
af af
x+y=f
ax ay
a a
x ax (e") + y ay (e" ) = e"
au au
or xe"  + yeO  = e"
ax ay
au au
or x  + y:; = 1 Hence Proved.
ax oy
.,' f = ell
Example 8 : If u ~ sin' { fx : Jy ) , show that
au au 1
x+y=tan u
ax ay 2
Solution: Here u = sin
l
1 x + y )
JX+fy
x+y
=> sin u = = {(say)
JX+fy
(U.P.T.U.2009)
Here f is a homogeneous function in x and y of degree ( 1  ~ ) i.e ~
:. By Euler's theorem for f, we have
26
Partial Di[ferentiation
af af 1
x+y=f
ax ay 2
a (. ) a (. ) 1.
or x  sm u + y  sin u =  sin u
ax ay 2
:. f = sin u
au au 1.
or x cos u  + Y cos u  =  sin u
ax ay 2
au au 1
or x+y=tanu.
ax ay 2
Hence Proved
Example 9: If u = tanI ,then prove that
(
X2 +y2)
x+y
au au 1.
x+y=sm2u
ax ay 2
(U.P.P.c.s. 2005)
x2 +y2
Solution: Here tan u = = f (say)
x+y
x2 +y2
Then for is a homogeneous function in x and y of degree 21 i.e 1.
x+y
:. By Euler's theorem for f, we have
af af
x+y=l.f
ax ay
a a
orx (tan u) + y (tan u) = tan u
ax ay
.,' f= tan u
au au
or x sec
2
u  + y sec
2
u  = tan u
ox oy
au au tan u. 1 .
or x  + y  = = sin u cos u =  SIn2U.
oX oy sec
2
u 2
Hence Proved
Example 10 : If u be a homogeneous function of degree n, then prove that
02U 02U ou
(i) x+y=(nl)
ox
2
oxoy ox
a
2
u a
2
u au
(ii) x+y=(nl)
oxoy oy2 oy
27
A Textbook of Engineering Mathematics Volume  I
a
2
u a
2
u a
2
u
(iii) x
2
2 + 2xy + y2 2 = n(n 1)u
ax axay ay
(U.P.T.U. 2005; Uttarakhand T. U. 2006)
Solution: Since u is a homogenous function of degree n, therefore by Euler's
theorem
au au
x+y=nu ............ (1)
ax ay
Differentiating (i) partially w.r.t. x, we get
x a 2 ~ + au .1+y a
2
u =n au
ax ax axdy ax
a
2
u a
2
u au au
or x+y=n
ax
2
axay ax ax
a
2
u a
2
u au
or x
2
+ y = (n 1) ........... (2)
ax axay ax
which prove the result (i)
Now differentiating (i) partially w.r.t. y, we get
x a
2
u + y a
2
u + 1. au = n au
ayax ay2 ay ay
a
2
u a
2
u au au
or x+y=n
ayax ay2 ay ay
a
2
u a
2
u au
orx+y2 =(n1) ................ (3)
ayax ay ay
Which proves the result (ii)
Multiplying (2) by x and (3) by Y and then adding, we get
a
2
u a
2
u a
2
u 2 a
2
u {au au}
x+xy+xy+y =(n1) x+y
ax
2
axay ayax ay2 ax ay
2 a
2
u a
2
u 2 a
2
u
or x +2xy+y =(n1) nu
ax
2
axay ay2
which proves the result (iii). Hence Proved
Example 11 : If u = xfJ ( ~ ) + f2 ( ~ ) , prove that
2 a
2
u a
2
u 2 a
2
u
x +2xy+y =0
ax
2
axay ay2
(LA.S. 2006; U.P.P.C.S. 1997; Uttarkhand T.U. 2006)
28
Partial Di[ferentiation
Solution: Let v = Xfl ( ~ ) and w = 6 ( ~ )
Then u = v + w .............. (i)
Now v = X f1 (y Ix) is a homogeneous function in x and y of degree one and we
prove that if u is a homogeneous function of x and y of degree n then
2 02U a
2
u 2 02U
x +2xy+y =n(nl) u
ax
2
oxoy ay2
(see Example 11 (iii))
:. Here we have
2 a
2
v a
2
v 2 (fv ..
x 2 +2xy+y 2 =1(11)v=0 ...................... (U)
ax axay ay
Also w = 6(y Ix) is a homogeneous function of x and y of degree zero so we have
2 a
2
w 2 a
2
w 2 a
2
w 0(0 1) 0 ( ... )
X 2 + xy + Y 2 =  W = ................ lU
ox axay oy
Adding (ii) and (iii) we have
a
2
a
2
a
2
X22 (u+v)+2xy(V+W)+y2_2 (v+w)=O
ax axay ay
2 a
2
u 02U 2 a
2
u
or x +2xy+y =0
ox
2
axay ay2
.: from (i) we have u = v + w. Hence Proved
Example 12 : If u = sin
1
[J + 2y + 3z l' show that
X
S
+ yS +ZS
au au au
x+y+z+3tanu =0
ax ay oz
(U.P.T.U. 2003)
. . . [x + 2y +3z 1
SolutIon: Here gIven u = sm
I
~
X
S
+ yS + ZS
x+2y+3z
~ sin u = = f (say)
~ x s +yS +Z8
Now here f is a homogeneous function in x, y, z of degree (1  4) i.e 3.
Hence, by Euler's theorem
af of of
x+y+z=3f
ax oy az
29
A Textbook of Engineering Mathematics Volume  I
or x ~ (sin u)+y ~ (sin u)+z ~ (sin u) = 3 sin u
ax oy oz
:. f = sin u
au au au.
or x cos u  + Y cos u  + Z cos U  = 3sm u
ax ay az
au au au sinu
or x+y+z=3
ax oy az cosu
au au au
orx+ y;+z+3tan u = o.
ax oy oz
Hence Proved
Example 13 : If u(x,y,z) = log (tan x + tan y + tan z) Prove that
. 2 au . 2 au . /'j au 2
sm x+sm y+sm.z=
ax ay az
Solution: we have
u(x,y,z) = log (tan x + tan y + tan z) ................. (i)
Differentiating (i) w.r.t. 'x' partially, we get
au = sec
2
x ..................... (ii)
ax tan x + tan y + tan z
Differentiating (i) w.r.t. 'y' partially we get
~ = tanx/t::;+tanz .................. (iii)
Again differentiating (i) w.r.t 'z' partially we get
au sec
2
z .
= ................ (IV)
az tan x + tan y + tan z
(U.P.T.U. 2006)
Multiplying (ii), (iii) and (iv) by sin 2x, sin 2y and sin 2z respectively and adding
them, we get
. au . 2 au . 2 au sin2x sec
2
x+sin2y sec
2
y+sin2z sec
2
z
sm2x +sm y +Sln z ="=
ax oy OZ tanx+tany+tanz
2sin x cosx.sec
2
x + 2sin ycos y.sec
2
y + 2sinz cosz.sec
2
z
=
tan x + tan y + tan z
2(tan x + tan y + tanz)
= ~    ~   ~
tan x + tany + tanz
=2
. 2 au . 2 au . 2 au 2 H P d
==>sm x+sm y+sm Z=. ence rove
ax oy az
30
Partial Differentiation
Total Differentiation
Introduction : In partial differentiation of a function of two or' more variables,
only one variable varies, But in total differentiation, increments are given in all
the variables,
Total differential Coefficient: If u = f(x,y)
where x = <I>(t), and y = \f'(t) then we can find the value of u in terms of t by
substituting from the last two equations in the first equation, Hence we can
regard u as a function of the single variable t, and find the ordinary differential
ffi
' du
coe Clent,
dt
Then du is called the total differential coefficient of u, to distinguish it from the
dt
, I d'fL 'I ffi' au d au
partia 1 lerentia coe Clent  an ,
ax ay
The problem now is to find du without actually substituting the values of x and y
dt
in f(x,y), we can obtain the requisite formula as follows:
Let (t + t) = X + h, "'( t + .) = y + k
Then by definition
du r f(x+h,y+k)f(x,y)
dt = 't
=lim o{f(X+h,y+k)f(x,y+k),h +
h 't k 't
, h dx ,k dy
Now hm =and hm =
't dt 't dt
Also, if k did not depend on h
I
, f(x + h, y + k)  f(x, Y + k)
h
would have been equal to
af(x,y + k)
ax
by definition
Moreover, supposing that 8f(x,y)/ Ox is a continuous function of y
lim af(x, y + k) = af(x, y)
ax ax
we shall assume, therefore that
I
, f(x+h,y+k)f(x,y+k) _ af(x,y)
h  ax
31
A Textbook o(Engineering Mathematics Volume  I
Hence
du af dx af dy
=+
dt ax dt ay dt
. du au dx au dy
Ie =+
. . dt ax . dt ay dt
Similarly, if u = f(XI' X2 ......... xn) and Xl, X2 ................ , Xn are all functions of t, we
can prove that
du au dX
1
au dX
2
au dXn
=.+.+ .................. +.
dt aX
1
dt aX
2
dt aXn dt
An important case: By supposing t to be the same, as x in the formula for two
variables, we get the following proposition:
When f(x,y) is a function of x and y, and y is a function of x, the total (i.e., the
ordinary) differential coefficient of f with respect to x is given by
df af af dy
=+.
dx ax ay dx
Now, if we have an implicit relation between x and y of the form f(x,y) = C
where C is a constant and y is a function of x, the above formula becomes
0= af +i!...dy
ax ay dx
Which gives the important formula
dy =_ af / ~
dx ax ay
Again, if f is a function of n variables Xl, X2, X3, .......... Xn, and X2, X3 ......... Xn are all
functions of Xl, the total (i.e. the ordinary) differential coefficient of f with respect
to Xl is given by
df af af dx
z
af dX3 af dX
n
=+.+.+ ..................... +.
dX
1
ax, aX
2
dx, aX
3
dx, aX
n
dX
1
Example 14 : If u = X log xy, where X
3
+y3+3xy = 1, find du .
dx
(V.P.T.V. 2002, 05)
Solution: Given u = X log xy ................. (i)
du au au dy ..
we know  =  + ................... (11)
dx ax ay dx
Now from (i) au = x . ~ y + log x y
ax xy
= 1+ log x Y
32
Partial Diiferentiation
dU 1 x
and =xx=
dy xy y
Again, we are given x
3
+y3+3xy =1, whence differentiating, we get
3x2 +3y2 dy + 3(X dy + y.1) = 0
dx dx
dy (x
2
+ y)
or  =  '::=.'
dx (y2 +x)
Substituting these values in (ii) we get
du x[ (X2 +y)]
=(l+logxy)+ (2 ) Answer.
dx y y +x
Example 15: If f(x, y) =0, <I>(y, z) =0 show that
df d<l> dz df d<l>
=
dy dZ dx dX dy
Solution: If f(x, y) =0 then dy = _(df ............. (i)
dx dX dy
if <I>(y, z) =0, then :; = ( )/( ..................... (ii)
Multiplying (i) and (ii), we have
dy. dz
dx dy dX dy dy dz
or = I (d<l> J. Hence Proved
dy dZ dx dX dy
Example 16 : If the curves f(x,y) =0 and <I>(x, y) =0 touch, show that at the point of
contact df. d<l> d<l> = 0
dX dy dy dX
Solution: For the curve f(x, y) =0, we have
dy dy =_(d<l/(d<l
dx dX dy dx dX dy
Also if two curves touch each other at a point then at that point the values of
(dy I dx) for the two curves must be the same,
Hence at the point of contact
33
A Textbook of Engineering Mathematics Volume  I
or )  ( ) = O. Hence Proved
Example 17: If $(x,y,z) =0 show that
1 =1
(U.P.T.U.2004)
Solution : The given relation defines y as a function of x and z. treating x as
constant
( J. =  ................. (i)
The given relation defines z as a function of x and y. Treating y as constant
(
az) = ......... (ii)
ax y $/az
S
'1 1 (ax) a$/ay ( ... )
Iml ar y, az z =  a$/ax .............. 111
Multiplying (i), (ii) and (iii) we get
(
ay) (az) (ax) _ 1 Hence Proved.
az x ax \' ay z
Change of Variables: If u is a function of x, y and x, yare functions of t and r,
then u is called a composite function of t and r.
Let u =f(x, y) and x = g(t, r), y = h(t, r) then the continuous first order partial
derivatives are
au au ax au ay
=+
at ax'at ay'at
au au ax au ay
=+
ar ax'ar ay'ar
Example18:lfu=u , show that X2_+y ;+z ;=0
(
y  x z  x) au 2 au 2 au
xy xz ax uy oZ
(U.P.T.U 2005)
(
yX zx)
Solution: Here given u = u ,
xy xz
= u (r, s)
yx zx
where r=and s=
xy zx
34
Partial Di[ferentintioll
1 1 1 1 .
=>r=and s= ................ (I)
x y x z
we know that
au au ar au as
=+
ax ar . ax as ax
1 1
= ~ ~ (  : 2 ) + ~ ~ (  :2)
:r=
x y
1 au 1 au
or x
2
au = _ au _ au ................... (ii)
ax ar as
.. au au ar au as
SImIlarly  = .+ .
ay ar ay as ay
from (i)
2 au au ( ... )
or y ay = ar .................... 111
and au=au.Or +au.as
az ar az as az
= au .0+ aU.2
ar as Z2
=> Z2 au = au ....................... (iv)
az as
Adding (i) (ii) and (iii) we get
2 au 2 au 2 au 0
x +y +z =
ax ay az
ar 1
=>=
ax x
2
1 1
:s=
x z
as 1
=>=
ax x
2
Hence Proved.
Example 19 : If u = u (y  z, z  x, x  y) Prove that
au + au + au =0
ax ay az
Solution: Here given u = u(y  z, z  x, x  y)
Let X = Y  z, Y :::0 Z  x and Z = x  y ................ (i)
Then u = u (X,Y,Z), where X, Y, Z are function of x,y and z.
Then
35
A Textbook of Engineering Mathematics Volume  I
au au ax au ay au az ..
++ (11)
ax  ax ax ayax az ax ................ .
au au ax au ay au az ...
=++ ................ (lll)
ay ax ay ay ay az ay
au au ax au ay au az .
and =++............ (lV)
az ax az ay az az az
with the help of (i), equations (ii), (iii) and (iv) gives.
au au au au au au
ax = ax 0+ ay (1)+ az (1) =  ay + az ................... (v)
au au au au au au .
ay = ax
1
+ aYO+ az(l)= ax  az .................... (Vl)
au au au au au au ..
and az = ax (1) + ay (1) + az (0) =  ax + ay ............... (V11)
Adding (v), (vi) and (vii) we get au + au + au = O. Hence Proved.
ax ay az
Example 20 : If z is a function of x and y and x = e
U
+ e
V
, y = e
u
 e
v
az az az az
Prove that =xy
au av ax ay
(AV.UP 2005)
Solution: Here z is a function of x and y, where x and yare functions of u and v.
:. az = az ax + az ay ................ (i)
au ax au ayau
az az ax az ay ..
and =.+............ (11)
av ax av ay av
Also given that
x =e
u
+ e
v
and y = e
u
e
V
ax u ax _y dy u ay y
:. au = e , av = e 'au = e 'av =e
:.From (i) we get
az = aZ(eU)+ az(_eU) ................ (iii)
au ax ay
and from (ii) we get
az = az (e
Y
)+ az (_e
v
) (iv)
av ax ay
Subtracting (iv) from (iii) we get
36
Partial Differentiation
az az (u v) az (u v) az
= e +e  e e 
au av ax ay
az az
=xy.
ax ay
Hence Proved.
Example 21 : If V = f(2x 3y, 3y 4z, 4z 2x), compute the value of
6V
x
+4Vy +3Vz.
Solution: Here given V = f(2x 3y, 3y 4z, 4z 2x)
Let X = 2x3y, Y =3y4z and Z = 4z2x ............. (i)
Then u = f(X, Y, Z), where X, Y, Z are function of x, y and z.
av av ax av ay av az ..
Then Vx ==++ ................ (11)
ax ax ax ay ax az ax
av av ax av ay av az .. ,
Vy ==++...................... (111)
. ay ax ay ayay az ay
av av ax av ay av az .
and V
z
==++ .................. (IV)
az ax az ay az az az
with the help of (i), equations (ii), (iii) and (iv) gives
V = av (2)+ av (0)+ av (2)
x ax ay az
or V = 2 (av _ av)
x ax az
=> 6V
x
= 12( ~ ~ ~ ~ ) .................. (v)
Now V = av (3) + av (3) + av (0)
Y ax ay az
or V =3( av + av)
y ax ay
=> 4V, = 12( ~ ~ + ~ ~ ) .................. (vi)
av av av
and V
7
= ax(O)+ ay(4)+ az(4)
or V =4( av + av)
z ay az
=> 3V
z
= 12(  ~ ~ + ~ ~ ) ................. (vii)
Adding (v), (vi) and (vii) we get
6V
x
+ 4Vy +3Vz =0 Answer.
37
(V.P.T.V.2009)
A Textbook Mathematics Volume  I
Example 22 : If x +y = 2e
B
coscp and x y = 2 ie
B
sincp, Prove that
a
2
v a
2
v a
2
v
ae
2
+ Ocp2 = 4xy oxoy
Solution: we have x+y = 2e
B
coscp, xy =2e
9
i sincp
=> x = e
B
(coscp + sincp) = eB. ei<l>
or x = e
B
+ it!>
and y = eB(coscp  i sincp) = eB. e i<p
or y = eB.it!>
since we know
av oV ox oV oy
=+
ae ox oe oy oe
oV O+i<p oV Ai<l>
=.e +.e
ox oy
oV oV
=x+y
ox oy
== x + y ................. (i)
oe ox oy
and oV = oV ox + oV oy
ocp ox ocp oy ocp
oV . 0+1<1> oV ( 1)' 01<1>
= a;.le + ay  .le
. oV . av
=lXly
ax ay
a . a . a ( .. )
................ 11
a<j> ax ay
A
. a
2
v a
2
v (a . a )(av . av)
gam oe2 + acp2 = ae + 1 acp ae 1 aq;
(V.P.T.V. 2001; V.P.P.C.S. 1997)
= av +y av +x av _y av)
ax ay ox ay ax oy ax ay
using (i) and (ii)
=2y
= 4
Y
[X a
2
v + .0]
ayax oX
38
Partial Differentiation
'iiv
=4xy axOy'
Hence Proved.
Example 23 : Transform the equation + = 0 into polar Coordinates
ax oy
(A.U.U.P. 2003, Delhi college of Engg. 2004, GG.S.I.P.U. 2004)
Solution: The relation Connecting Cartesian Coordinates (x, y) with polar Co
ordinates (r, 0) are
x = r cosO, y = r sinO
Squaring and adding r2 = x
2
+y2
tan 0 = r
x
ar = x = r cos 0 = cos 0, ar = y = r sin
0
= sin 0
ax + y2 r ay + y2 r
ao 1 ( y) y sinO
ax = 1 + y2 7 =  x
2
+ y2 = r
x
2
ao 1 1 x cosO
and =.=? ? =
ay 1 + y2 X x + y r
x
2
Here u is a composite function of x and y
au = au ar + au ao = cos 0 au _ sin 0 au
ax ar ax ao ax . ar r ao
a a sinO a .
 == cosO  ......................... (1)
ax ar r ao
Also au = au . ar + au . ao
ay ar ay ao ay
. oau cosO au
=sln +
ar r ao
cosO ................. (ii)
ay ar r ao
Now we shall make use of the equivalence of Cartesian & polar operators as
given by (i) and (ii)
a
2
u = = (cosoi. _ sinO au _ sinO au)
ax
2
ax ax ar r ao ar r ao
39
A Textbook of Engineering Mathematics Volume  I
S
a ( S au sin S au) sin S a (' S au sin S au)
=cos  cos   cos 
ar ar r as r as ar r as
= COSs[COSS a
2
u _ sinS au _ sinS a
2
u ]_ sins[sins au + cosS a
2
u _ coss au _ sinS a
2
u]
ar2 r as r2 r aras r ar asar r as r as
2
_ 2 S a
2
u 2cosSsinS au sin
2
S au _ 2cosSsinS a
2
u sin
2
S a
2
u ('"
 cos ar2 + r2 as + r ar r aras + r2 as
2
...................... 111)
And a
2
u = J = case au + case au)
ay2 ay ay ar r ae ar r ae
, e a ( , e au cos e au ) cos e a ( , e au cos e au)
=SIn  SIn + + sm +
 ar ar r ae r ae ar r ae
, e[ , e
a2u
case au case a
2
u] cose[ e
au
'e a
2
u sine au case a
2
u]
=sIn SIn + + cos +sm +
ar2 r2 ae r arae r ar aear r ae r ae
2
2 ea2u 2cosesinS au cos
2
e au 2cosSsinS a
2
u cos
2
S a
2
u (' )
= sin ar2  r2 as + rih + r arae + ras
2
.......... ' IV
Adding (iii) and (iv)
a
2
u a
2
u a
2
u 1 au 1 a
2
u
+=++
ax
2
ayZ ar
z
r ar rZ ae
z
azu azu ,azu 1 au 1 a
2
u
Therefore  +  = 0 transforms mto  + + = 0
ax
z
ayZ ar2 r ar rZ ae
2
Exercise:
1. If u = tan
1
xy , show that
+xz +yZ
azu 1
axay (l+xz +y2t2
2, If u(x+y) = x
2
+ y2, prove that
(
au _ au)Z = 4 (1 _ au _ au)
ax ay ax ay
3, If u = x
2
tan
l
( : )  y2 tan
1
) ,prove that
a2u XZ _ y2
= ".,,
ayax x
2
+yZ
40
Partial Di[ferentiation
(V.P.T.V. Special Exam 2001)
4. If z = f(x+ay) +<j>(x ay), prove that
a
2
z 2 a
2
z
=a
ay2 ax
2
a
2
u a
2
u
5. If u = 2(ax+by)2  (X2+y2) and a
2
+b
2
=1, find the value of
ax2
+ ay2 .
6. If Y =(x2+y2+z2)1/2, Show that
a
2
v a
2
y a
2
y
++=0
ax
2
ay2 az
2
7. If u = (X2+y2 +Z2)1/2, then prove that
a
2
u a
2
u a
2
u 2
++=
ax
2
ay2 az
2
u
Ans=O
(P.T.V.1999)
r2
9.lf S = t
n
e 4t ,find the value of n which will make ~ ~ ( r 2 as) = as
10. If x = r cosS, y = r sinS, prove that
(i) ar = ax (ii) .!. ax = r as
ax ar r as ax
az az
11. If z = eax+by f(ax  by), Show that b+ a = 2abz
ax ay
rar ar at
3
Ans. n=
2
2
12. If Y = f(r) and r2 = x2+y2 +z2, Prove that Y xx+ Yyy + Y zz = f"(r) +  f'(r)
r
13. Find p and q, if x = fa (sin u + cos v), y = fa (cos u  sin v), z =l+sin(u  v)
az az .
where p and q means  and . respectively.
ax ay
[ Hint. X2+y2 =2az
x2 +y2
:. z = ]
2a
Ans. p = ~ , q = r
a a
41
A Textbook Mathematics Volume  I
au au
14. If u =(12xy +y2)1/2, Prove that x _ y_ = y
2
u
3
ax ay
(M.D.U.2001)
15. Verify Euler's theorem for the function z = sin
1
+tan
1
r
y x
(Delhi college of Engg., March 2004)
x
2
+ 2 a a
16. If u =sin
1
y, show that + = tan u
x+y ax ay
(B.P.s.C 1997; U.P.P.CS. 1990; J.N.T.U. 1999, G.GS.I.P.U. 2007, M.D.U. 2002,03
V.T.U.2003)
x
3
+ y3 az az .
17. If z =tan
1
,Prove that x + y = sm2z
xy ay
(U.P.T.U. 2001, M.D.U. 2002, GGS.I.P.U. 2006)
[
X4 + y4 ) au au
18. If u = log ,show that x + y = 3
x+y ay
(U.P.T.U.2001)
x+ y au au 1
19. If u = cos
1
C r:;' show that x + y + cotu = 0
"X + '1/ Y ax ay 2
(G.G.S.I.P.U. 2006, V.T.U. 2004)
[
X
3
+
l
) a2u a
2
u a
2
u .
20. If u = tan
1
,show that x
2
2 + 2xy+ y2 = 2cos3u sm u
x  y ax axay v J
(U.P.P.CS.1993)
(A.U.U.P. 2005, P.T.U. 2002, Delhi college of Engg 2004.)
x+y
21. If u = sin
1
c r:; , Prove that
"X + 'l/y
. au au 1
(1) x+y+tanu
ax ay 2
.. ) 2 a
2
u a
2
u 2 a
2
u sin u cos2u
(11 x +2xy+y =
ax
2
axay ay2 4 cos
3
u
(A.U.U.P. 2005, M.D.U. 2000, 2004, Delhi college of Engg. 2005)
22. If u = sin1 f x:;: + }1/2 ,then show that
lx +y
a
2
u a
2
u a
2
u tan u
x
2
2 + 2xy + y2 2 = (13 + tan
2
u)
ax axay ay 144
(Delhi college of Engg. 2004, M.D.U. 2001, 2003)
42
Partial Differentiation
1 1 log x  log y af af
23. Iff(x,y)= 2 ++ 2 2 ,showthat
x xy x +y ux uy
(PTV 2004)
24. If z = X4 y2 sin
1
+ log x log Y show that x az + y az = 6X
4
y
2 sin
1
y ax ay y
(V.P.T.V.2003)
25. Show that x:l + Y  + z:l = 2 tan u where u = sin
1
=
uX ay uZ ax + by + cz
(V.P.T.V.2003)
j
1t(2X
2
+ y2 + xzt2 )
26. If V = loge sin 1/3 Prove that when x =0, y =1, z =2
2(X2 +xy+2YZ+Z2)
av av av 1t
x+y+z=
ax ay az 12
(V.P.P.C.S.1991)
27. If u = x
2
 y2 + sin yz where y = ex and z = log x; find du
dx
1
Ans. 2(x  e
2x
)+e
x
cos (ex log x) (log x + ).
x
au au au
28. If u = f(r, s) and r = x+y, s = xy; show that  +  = 2
ax ay ar
02U a
2
u (a
2
u 02U)
29. If x = e
r
cosS, y = e
r
sinS, show that ax
2
+ ay2 = e
2r
ar2 + OS2
30. If z is a function of x and y and u and v be two other variables such that
u = Ix + my, v = Iy  mx, show that  +  = (12 + m
2
)  + 
a
2
z a
2
z (a
2
z a
2
z)
ax
2
ay2 ou
2
av
2
u az v az v
2
az
31. If z = f(x, y), where x = uv and y = , then show that  = . .
v ay 2 au 2u av
(A.U.U.P.2005)
32. If z = + y2 and x
3
+y3 +3axy = 5a
2
, find the value of dz when x = y =a.
dx
au au
33. If u = f(y Ix) show that x+ y = 0
ax ay
(U.P.P.C.S.1997)
43
A Textbook of Engineering Mathematics Volume  I
az
2
a
2
z
34. Prove that if z = <I> (y + ax) + ",(y  ax) then a
2
   = 0 for any twice
ay2 ax
2
differentiable <I> and "', a is a constant.
3S.1f u= ,I that x au + y au + z au = 0
y z x ax ay az
Tick the Correct answer from the choices given below
1. If u = f(y Ix), then
. au au
(1) xy=O
ax ay
(
") au au 0
11 x+y=
ax ay
(
"') au au
111 x+y=u
ax ay
. au au
(IV) xy=l
ax ay
03
U
3. If u = eXYz, then 0 is equal to
ox yoz
(i) (x
2
+y2+z2+3xyz)e
x
yz
(iii) (1 +x2+y2+z2)e
x
yz
(ii) (1 +x2y2z2+3xyz)e
x
yz
(iv) (3x+3y+3z+x2y2z2)e
x
yz
au au au.
4. If u = loge (X
3
+y3+z3 3xyz), then  +  +  IS equal to
ax ay az
(i).!. (x+y+z) (ii) .!. (x+y+z)
2 3
(
"') 2
111
x+y+z
(iv) 3
x+y+z
oz az.
S.1f z = xy f(y Ix), then x + yIS equal to
ax ay
(i) z (ii) 2z
44
(LAS. 2007)
(LAS. 1996)
Ans. (ii)
(LAS. 1999)
Ans. (iii)
(LAS. 1999)
Ans. (ii)
Ans. (iv)
(iii) xz (iv)yz
Partial Differentiation
Ans. (ii)
(U.P.P.CS.1994)
6. Sino! ( ~ ) is a homogeneous function of x, y of degree
(i) 1 (ii) 2
(iii) 3 (iv) 0
7. If z = uv
u
2
+ v
2
 xy =0
u
2
 v
2
+ 3x + Y =0
Then az is equal to
ax
(i) u + v
3
2 2
(
.. ) U + V
111
2uv
2
22
(ii) u v
2uv
u
2
_ 3v
2
(iv) 
2uv
8.lfu=sin! Y,thenx+yis
(
X2+ 2) af af
x+ y ax ay
(i) f (ii) 2f
(iii) tan f (iv) sin f
9. If u = X4 y2, where x = t
2
and y = P, then du is
dt
(i) 22t
13
(
... ) 22 til
111 
23
(ii) 14 t22
(iv) 14t
13
(
X3 + y3 ) au au .
10. If u = tan! the value of x  + y  IS
X  Y ax ay
(i) tan 2u (ii) cos 2u
(iii) sin 2u (iv) sec
2
2u
45
Ans. (iv)
(U.P.P.CS. 1994,1995)
Ans. (ii)
(I. AS. 1994)
Ans. (iii)
(LAS. 1994, U.P.P.CS. 1994)
Ans. (iv)
(RAS 1993)
Ans. (iii)
A Textbook o(Engineering Mathematics Volume  I
(R.A.S.1994)
11. If XX yy ZZ = c, then a
2
z =  (x log ex) n when (x = y = z) if n is equal to
axay
(i) !
2
(iii) 1
(
") 1
11
3
(iv) 2
12. If u = x log xy, where x
3
+y3 +3xy =1, then du is equal to
dx
X (X2 + y)
(i) (1 + log xy)  2
Y Y +x
(iii) (Ilog xy) _ ~ ( X 2 + y)
Y y2 +x
(ii) (1 + log xy) _r( y2 + x)
X x
2
+y
. Y (y2 + x)
(IV) (1 log xy)  
X x
2
+y
(
X2 + y2 ) au au .
13. If u = log Then x + y IS equal to
x+y ax oy
(i) 0 (ii) 1
(iii) u (iv) eu
. X
1/2
_y1/2 2 a
2
u 02U 2 a
2
u
14. Ifu = sm
1
1/2 1/2 ,thenx 2 +2xy+ y 2
x + y ax axay ay
is equal to
(i) 0 (
") U
11 
4
(
"') u
111 
4
(
. ) 3u
IV 
4
~ 2 2 au au au.
IS.Ifu= x +y2+Z ,then x+y+zIsequalto
ax oy az
(i) 0 (ii) u
(
"') 1 (. ) U
111  IV 
2u 2
46
Ans. (iii)
(V.P.P.C.S.1994)
Ans. (i)
Ans. (ii)
Ans. (i)
Ans. (ii)
Partial Di[ferentiation
x
2
y2 Z2
16.lfu = x Y z ,then Ux +u
y
+ U
Z
is equal to
1 1 1
(i) 0 (ii) 1
(iii) x + Y + Z (iv) 2(x + Y + z)
au au au.
17. If u = f(yz, zx, xy) then  +  +  IS equal to
ax ay az
(i) x + Y + z (ii) 1 + x + y + z
(iii) 1 (iv) 0
a
2
z a
2
z
18. If z = f(r) and r2 = x
2
+y2, then 2 + 2 is equal to
ax ay
(i) fll(r) + r f'(r) (ii) f'(r) + r fll(r)
(iii) f'(r) + ! fll(r) (iv) fll(r) + ! f'(r)
r r
Ans. (i)
Ans. (iv)
(U.P.P.C.S. 1994)
Ans. (iv)
19. If /+t+f=l, then (au)2 +(:)2 +(au)2 is equal to
a + u b + u c + u ax vy az
. au au au .. 1 (au au au )
(1) x+y+z (u)  x+y+z
ax ay az 2 ax oy oz
(iii) 2(X au + y au + z au) (iv) None of these
ax ay az
(i) r
m

2
(iii) m (m+1)r
m

2
(ii) m (m  1)rm2
(iv) (m2 1)r
m

2
Ans. (iii)
Ans. (iii)
21. If u is a homogeneous function of degree n in x, y, z and if u = f(x, y, z) where
x, y, z are the first derivatives of u with respect to x, y, z respectively then
af af af .
x+y+zlsequal to
ax ay az
(i)nu (ii)n(n1)u
47
A Textbook of Enzineerinz Mathematics Volume  I
(
... ) 1
111 u
nl
(iv) _n_u
n 1
22. The total differential dz of z which is a function of x and y is equal to
(
.) d d d ( .. ) d dz d dz d
1 z= x+ Y 11 Z= x+ Y
dx dy
(
... ) d aZ
d
az d
III Z= x+ y
ax ay
23. If f(x, y) =0, <I>(y, z) =0, then
(iv) dz = az + az
dx ax ay
(i) ~ . a<l> = ~ a<l>. dz (ii) ~ . a<l>. dz = af. a<l>
ay az ax ay dx ay az dx ax ax
(1
11) _af . a<l> . dz ___ af . a<l> (. ) N f th
Ivane a ese
ay az dx ax ay
24. If u = yx, then au is equal to
ax
(i) yx ( 2 + log y)
Y
(iii) yx log Y
(ii) xyxl
(iv) yx log x
25. A function f(x, y) is said to be homogenous of degree n in x, y if
(i) f(tx, ty) = t2n f(x, y)
(ii) f(tx, ty) = t
n

1
f(x, y)
(iii) t is of the form xn f(x/y)
(iv) t is of the form xn f(y I x)
26. Consider the Assertion (A) and Reason (R) given below:
au au
Assertion (A)  If u = xy f(y Ix), then x+ y = 2u
ax ay
Reason (R)  Given function u is homogenous of degree 2 in x and y.
of these statements 
(i) Both A and R are true and R is the correct explanation of A.
(ii) Both A and R are true and R is not the correct explanation of A.
(iii) A is true but R is false.
(iv) A is false but R is true.
27. Consider the following statements:
48
Ans. (i)
Ans. (iii)
Ans. (iii)
Ans. (iii)
Ans. (iv)
Ans. (i)
Partial Oitferentiatioll
. . ax 1 ax 1
Assertion (A)  for x = r cosS, y = r smS  = ":I and:; = ":I
ar ar / uX uS uS / ax
dy 1
Reason (R)  = 
dx dx/ dy
of these statement
(i) Both A and R are true and R is the correct explanation of A.
(ii) Both A and R are true and R is not the correct explanation of A.
(iii) A is true but R is false.
(iv) A is false but R is true.
ax ay
Ans. (iv)
(LA.S. 1995)
(U.P.P.CS.1994)
. ar ar
28. If x = r cosS, y = r smS, then ":I is
ax uy
(i) r
C) 1
11 
r
(iii) 2r (iv) 2
2r
29. Match the list I with II
List I
as as
x
2
y a
2
u a
2
u
(a) If u = , then xa 2 + ya
x+y x xay
XI/2  yl/2 2 a
2
u a
2
u 2 a
2
u
(b) If u = 1/4 1/4 ' then x a 2 + 2xy a + y a 2
x +y x ax y y
a
2
u a
2
u a
2
u
(c) If u = X
1
/
2
+ yl/2 then x
2
+ 2xy+ y2_
ax
2
axay ay2
au au
(d) If u = f (y/x), thenx; + y
uX ay
List II
3
(p) u
16
(q) 2 au
ax
(r) 0
(s) u/4
49
Ans. (i)
A Textbook of Engineering Mathematics Volume 1
correct match is
a b c d
(i)
P
q r s
(ii) q
P
s r
(iii) q
P
r s
(iv)
P
q s r
au ou au
29. If u = 3x2yz + 2yz3 + 6 X4, thenx+ y + zequal to
ax oy az
(i) 6x
2
yz + 4xy3 + 12 X4
(ii) 3x2yz + 2yz3 + 6x
4
(iii) 12 x2yz + 6yz3 + 12 X4
(iv) 12 x2yz + Byz3 + 24 X4
o2f
30. If f(x, y) = X4+X2y2+y4, then is equal to
oxay
(i) 4xy (ii) 12 x
2
+2y2
(iii) 2x
2
+ 12 y2 (iv) 4x
3
+2xy2
50
Ans. (ii)
(U.P.P.C.S.1995)
Ans. (iv)
(u.P.P.c.s 1994)
Ans. (i)
Rules for tracing Cartesian Curves :
Chapter 3
Curve Tracing
I. Symmetry  The most important point while tracing the curves is to judge its
symmetry which we do as following:
(a) If the equation of the curve involves even and only even powers of x, then
there is symmetry about y axis. The reason is that if we obtain a certain value of y
by putting x =a in the equation then the same value of y will be obtained by
putting x = a in the equation because it contains even and only even powers in x
e.g. the parabola x
2
= 4ay and the circle X2+y2 =a
2
are both symmetrical about y
axis.
Note. The stress on the words 'even and only even' should be observed. x
2
+y2 =
ax is not symmetrical about y axis as it involves odd powers of x as well.
(b) Similarly, if the equation of the curve involves even and only even powers of
2 2
Y then the curve is symmetrical about x axis, e.g. ellipse;' + y 2 = 1 and parabola
a b
y2 = 4ax are symmetrical about x axis.
(c) If the equation of the curve involves even and only even powers of x as well as
of y, then the curve is symmetrical about both the axes i.e. the circle x
2
+y2 = a
2
or
2 2
ellipse;' + ~ = 1 are both symmetrical about the axes.
a b
(d) If x be changed into x and y be changed into y, and the equation of the
carve remains unchanged, then the there is symmetry in opposite quadrants e.g.
xy =a
2
is the equation of the hyperbola referred to asympototes as axes which is
symmetrical in 1st and 3rd quadrants.
(e) If x be changed into y and y into x and the equation of the curve remains
unchanged, then curve is symmetrical about the line y = x i.e. a line passing
through origin and making an angle 45
0
with positive direction of xaxis.
Note. In this case solve the equation of the curve with the line y =x and find the
point of intersection e.g. x
3
+y3 = 3axy is symmetrical about y =x and cuts it in
points (0, 0) and ( ~ a , 3;).
51
A Textbook o[Engineering Mathematics Volume  I
II Points
We should find points where the curve cuts the axes which are obtained as
follows;
Put y = in the equation of the curve and solve for x and thus you will find the
2 2
points where the curve cuts the x axis e.g. \ + y 2 = 1 putting y =0 we get x
2
=a
2
.
a b
:. x =a and a, and hence ellipse cuts the x axis in points (a, 0) and (a, 0)
Similarly, put x =0 in the equation of the curve and solve for y and thus you will
2 2
find the points where the curve cuts the y axis e.g. ;. + Y2 = 1 putting x =0, we
a b
get y =b and b and hence the ellipse cuts the y axis in points (0, b) and (0, b).
Again y2 = 4ax putting x =0, we get y =0 and it we put y =0 we get x =0 and hence
the point (0, 0) i.e. origin lies on the curve. The best way for detecting whether the
origin lies on the curve is to see that the equation does not contain any constant
term just as y2 = 4ax contains no constant term and hence it passes through the
origin. The circle x
2
+y2 =a
2
contains the constant term a
2
and hence it does not
pass through the origin.
III. Tangents: After we have found the points which lies on the curve we shall
try to find tangents at those points.
(a) Tangents at origin: In case we find that origin is a point on the curve, then
the tangents at the origin are obtained by equating to zero the lowest degree
terms occurring in the equation of the curve e.g. y2 = 4ax passes through origin
and the lowest degree terms occurring in it is 4 ax which when equated to zero
gives x =0 i.e. y axis which is tangent to the parabola at the origin.
Again the equation a
2
y2 = a
2
x
2
 X4 contains no constant term and as such passes
through the origin and the total lowest degree terms brought on one side are a
2
y2
a
2
x
2
which when equated to zero y2 x
2
=0 or Y = x which will be tangents at
the origin to the curve i.e. the curve shall be touching both these two lines at the
origin.
52
Curve Tracing
y
y =x y=x
x '           ~ ~ ~           x
Y'
Again x
3
+y3 = 3axy passes through the origin and the lowest degree term is 3axy,
which when equated to zero gives x =0, y =0 as the tangents at the origin i.e.
curve will touch the axes at the origin.
Again y2(a+x) = x
2
(3ax) clearly passes through the origin and the total lowest
degree terms brought on one side are ay2 3ax
2
which when equated to zero gives
y2 3x
2
=0 or y = J3 x which will be tangents at the origin to the curve.
Note. We shall write tangents at the origin as To, n.
(b) Tangents at any other points which is not origin: Suppose there lines a point
(h, k) on the curve, then in order to find the tangent at (h, k) we find the value of
dy from the equation of the curve and substitute in it (h, k) which gives us the
dx
slope of the tangent at (h, k) and the equation of the tangent will be a line passing
through (h, k) and with the above slope. If the slope comes out to be zero then the
tangent will be parallel to x axis, e.g. we have seen before that x
3
+y3 = 3axy
passes throu
g
h(3a, 3a). Let us find dy by differentiating the equation of the
2 2 dx
curves w.r.t.x
3X2+3y2 dy =3a(y +x dy )
dx dx
53
A Textbook of Engineering Mathematics Volume  I
Y
p(3a 3a)
2' 2
X        = ' ~     ~ .......... X'
Y'
dy ay  x
2
d I b b . h . ( 3a 3a). I 1 d
:.  = 2 an Its va ue y su stituting t e pomt , IS equa to  an
dx y ax 2 2
hence the slope of the tangent at( 3
2
a , 3
2
a) is 1 i.e., it will make an angle of 1350.
IV. Asymptotes which are parallel to the axes
The asymptotes parallel to the x axis may be obtained by equating to zero the
coefficient of highest degree terms in x, provided it is not a constant.
Similarly, the asymptotes parallel to the axis of y may be obtained by equating to
zero the coefficient of highest degree terms in y, provided it is not constant.
V.Region
If for some value of x greater than some quantity say a the corresponding values
of y corne out to be imaginary, then no part of the curve will lie beyond x =a.
Similarly if for some value of y greater than some quantity b say the
corresponding values of x come out to be imaginary, then no part of the curve
shall lie beyond y =b. e.g. y2(2a x) =x
3
If x is greater than 2a, then 2a x will be negative and hence y2 will be negative.
Therefore y will be imaginary for values of x greater than 2a and as such no part
of this curve will lie beyond x =2a. Again if x is negative then 2ax is positive but
x
3
becomes negative and therefore y2 will be negative and as such y imaginary.
Hence no part of the curve shall exist in the negative side of x axis.
In case the equation of the curve can be arranged as a quadratic in y or x, then it
is convenient to find the values of one variable in terms of the other and discuss
the reality of the roots.
Note. The existence of loops is generally found by this.
54
Curve Tracinz
Example 1: Trace the curve y2(a x) =x
3
cissoid.
(V.P.T.U.2005)
Solution: (i) Since there are even and only even powers of y, symmetry is about
xaxis.
(ii) Putting x =0 we get y =0 and putting y =0, we get x =0 and hence (0, 0) is the
only point on the curve.
(iii) Tangents at origin are given by the lowest degree terms equated to zero i.e.
ay2 =0 :. y =0 Le. xaxis is tangent.
(iv) The equation is of 3rd degree and x
3
is present but is missing and hence
equate to zero the coefficient of y2 which is ax. Therefore x =a is an asymptote
parallel to y axis.
(v) If either x is negative or greater than a, then corresponding values of yare
imaginary. Therefore the curve is within the lines x =0 and x =a with the above
five points the shape of the curve is as shown in the figure given below.
Y
x=a
Asy
A
X' (a, 0)
Y'
Example 2 : Trace the curve y2(a+x) = x
2
(ax) (Strophoid)
or
(X2 + y2) X  a(x
2
 y2) =0
X
(Uttarakhand TU 2006, U.P.P.C.S. 1996, B.P.S.C 2007)
Solution: (i) Symmetry about x axis
(ii) Points y =0 then x =0 and a, :. (0, 0) and (a, 0) and when x =0 then y =0 :. (0,0)
Hence (0, 0) and (a, 0) are the only two points.
(iii) T(o, 0) are a(x2  y2) =0 i.e. y = x
(iv) The asymptotes is x + a =0 i.e. x = a, on equating to zero the coefficient of y2
as y3 is missing.
(v) Again if x is greater than a then the value of y2 is negative and hence y
imaginary. Therefore the curve does not go beyond x =a, similarly it does not go
beyond x = a.
With the above five points the shape of the curve is as shown in the figure given
below.
55
A Textbook of Enzineerinz Mathematics Volume  I
ASY
x =a
X
I   ~   ~   ~ ~     ~ ~  = ~  X
A' A(a, 0)
(a, 0)
Example 3: Trace the curve a
2
y2 = x2(a
L
x2)
or a
2
y2 =a
2
x
2
 X4
Solution: (i) Symmetry about both axes.
(ii) Points (0,0), (a, 0), (a, 0)
(iii) To,o are y = x and T(a, 0) is y axis and T(_ a, 0) is y axis.
(iv) No asymptotes
(v) x cannot be greater than a in magnitude on either side of x axis with the above
data the shape of the curve is as shown in figure.
Y
(a, 0) .;r. (a, 0)
yl
y =x
Example 4: Trace the curve
x
3
+y3 =3axy (Folium of Descartes)
(V.P.T.V.2003)
Solution: (i) If x be changed into y and y into x, the equation of the curve is
unaltered and hence the symmetry is about the line y =x.
(ii) (0, 0) is one point and the other point is obtained by solving the equation with
h
h' (3a 3a)
y =x w IC IS 2 '2 .
(iii) Too are x =0, y =0 and the value of dYat (3a, 3a) is 1, which shows that
, ili 2 2
tangent at that point makes an angle of 135
0
with x axis.
56
Curve Tracing
(iv) There is no asymptote which is parallel to either axis but there is an oblique
asymptote x + y +a =0.
(v) x and y both cannot be negative, because that will make L.H.5 of the curve
negative and R.H.S positive. This implies no portion of the curve lies in the third
quadrant.
Shape of the curve given below.
y
y = x or 8 = 450
XI
Example 5 : Trace the curve
a
2
x
2
= y3 (2a y)
X
Solution: (i) symmetry about yaxis as the power of x are even.
(ii) x =0 we get y =0 and 2a, :. (0, 0) and (0, 2a) are the points; when y =0 we get
x =0, :. (0,0)
(iii) T(O, 0) are given by a
2
x
2
=0 i.e. x=o i.e. yaxis.
(iv) No asymptotes.
(v) when y is either greater than 2a or negative, then X2 is negative and hence x
will be imaginary. Hence no part of curve lies below y =0 i.e. xaxis and beyond y
=2a with the above data the shape of the curve is as shown in figure given below.
XI
y
(0,2a) B
yl
X
Polar Coordinates: If we have any horizontal line OX called the initial line and
another line called the revolving line makes an angle 8 with the initial line then
the polar coordinates of a point P on it where OP = rare (r, 8).
57
A Textbook of Engineering Mathematics Volume  I
P (r, 9)
OP=r
______________ x
o
Initial Line
The point 0 is called the pole and the angle 9 is called the vectorial angle of the
point P and the length r is called the radius vector.
Rules for tracing Polar Curves 
I. Symmetry
If in the given equation 9 be changed into 9 and the equation of the curve
remains unchanged, then there is symmetry about the initial line.
If the equation of the curve remains unchanged when r is changed into r, then
there is symmetry about the pole.
II. Plotting the points Give to 9 certain value and find the corresponding values
of r and then plot points. Sometimes it is inconvenient to find the corresponding
values of r for certain values of 9 and even if we find they involve radical, then in
such cases we should consider a particular region for 9 and as certain whether r
increases or decreases in that region.
F tho b h f II
or IS we must remem er t e o owmg:
At9=O 1t 1t 31t 21t
 
2 2
sin9= Min Max Min Max (in Magnitude) Min
cose = Max Min Max (in Magnitude) Min Max
e.g. r = a( 1 + cose)
III. Region No part of the curve shall exist for these values of e which make
corresponding values of r imaginary.
e.g. r2 =a
2
cosze
58
Curve Tracing
As 8 increases from 45
0
to 135
0
, then 28 increases from 90
0
to 270
0
and this interval
cos28 will always be negative and consequently r2 will also be negative, which
shall give imaginary values of r. Hence no part of the curve shall exist between 8
= 45
0
and 8 = 135
0
We also find the limits to the values of r.
i.e. r =a sin28
Now whatever 8 may be sin28 can never be greater than unity and hence a sin28
or shall never be greater then a i.e. the curve shall entirely lie within the circle r
=a.
Example 6 : Trace the curve
r = a (1 +cos8) (Cardiod)
Solution: If we change 8 into 8, the equation of the curve remains unchanged
and hence there is symmetry about the initial line.
8=0 60
0
90
0
120
0
1800
r = 2a 3a a a 0
 
2 2
(a,900) C ____ (_3a/2, 60)
(aj2, 12(0) 0
(2a, 0)
X '             ~ ~     ~   ~ r .          X
A
o
We observe that r continually goes on decreasing as 8 increases from 0
0
to 180
0
The above points trace the curve above the initial line and the other portion is
drawn by symmetry. With the above data the shape of the curve is as in the
figure above.
Example 7: Trace
r = a (1 cos8) (Cardiod)
Solution: Trace as above
59
A Textbook of Engineering Mathematics Volume  I
600)
A 2'
(2a, 180) (0, 0)
o
Example 8: Trace the curve r2 = a
2
cos 28
(Lemniscate of Bernoulli)
(V.P.T.V. 200809. 2001)
Solution: Symmetry about the initial line putting r = 0, cos 28 =0 or 28 = or
2
e = . Hence the straight lines 8 = are the tangents to the curve at the pole.
4 4
A8 fr Ot bl s vanes om o 1t, r vanes as glven eow:
8 =0
0
30
0
45
0
90
0
135
0
150
0
180
0
r2 =a
2
a
2
0 a
2
0
a
2
a
2
 
2 2
r = ia
0 imaginary 0
ia
J2, J2,
The above data shows that curve does not exist for values of 8 lying between 45
0
and 135. With the above data the shape of the curve is shown as given below.
Example 9: Trace the curve
60
Curve Tracing
r = a cos28
(V.P.T.V.2003)
Solution: Symmetry about the initial line putting r =0, cos28=0 or 28 = .:: or
2
8 = .:: . i.e. the straight lines 8 = .:: are the tangents to the curve at the pole.
4 4
C d' I fe d bel orrespon mg va ues 0 an rare Iven ow.
e = 0
0
30
0
45
0
60
0
90
0
120
0
135
0
150
0
1800
r=a a 0 a a a 0 a a
  

2 2 2 2
Plot these points and due to symmetry about initial line the other portion can be
traced.
7t
8= 
2
/
"
"
/
/
7t
e= 
,/ 4
e = 7t 8 = 0
e = 57t
4
,
,
,
,
,
,
" 77t
e=
4
Note: The curve is of the form of r = a cos n8 (or r = a sin n8) and in such case
there will be n or 2n equal loops according as n is odd or even.
Example 10 : Trace the curve
r = a cos38 (V.P.P.C.S. 2004)
Symmetry about the initial line. Putting r =0, we get cos38 =0
7t 37t 51t
or 38=+ + +
2'2'2
1t 1t 51t
or 8=+ + +
6'2' 6
Which give the tangents at the pole
dr . 38
','=aSln
de
61
A Textbook of Engineering Mathematics Volume  I
. dr . 38 0
:. equating  to zero we get sm =
or 38 =0,
or 8 = 0,
WhO h .
IC gIve
8 =0
r=a
d8
n,
n
"3'
2n,
2n
3'
3n,
n,
th e maXImum va ues 0
n n
 
6 3
0 a
f8
n
4n
4n
3
d an

2
0
are ~ lven b I e ow:
2n 5n
 
3 6
a 0
n
a
plot the above points and with the above data the shape of the curve is a shown
in figure given below:
1t
8 =
2 1t
8= 
, 3
, 1t
,," , 8 = 
, , ~ ", 6
, " , ,
, ...
e = 1t      = '  ~ ~    +    e = 0
Example 11: Trace the curve r = a sin 28 (U.P.T.U.2001)
Solution: Symmetry about the line 8 =!: putting r =0, sin28 =0
2
or 28 =0, n, 2n, 3n
n 3n
or 8 = 0, 2' ,n, 2
Which are the tangent at the pole.
dr . dr
 = 2acos28 , Equating  to zero, we get cos2e =0
d8 de
62
Curve Tracing
1t
8 =
2
"
"
1t
8= 
,/ 4
8 = 1t 8 = 0
8 = 51t
4
1t 31t 51t 71t
or 28=, , ,
2 2 2 2
1t 31t 51t 71t
or 8=, , , 
4 4 4 4
8 = 31t
2
,
,
" 71t
8=
4
Which gives the value of r and which is equal to a sin = = a.
Th d' al f 8 d b I e correspon mgv ueso an rare glVen eow.
8=0 1t 1t 31t 1t 51t 31t 71t
   
 
4 2 4 4 2 4
r=O a 0 a 0 a 0 a
21t
0
Plot the above points and with above data the shape of the curve is shown as
above.
Parametric Curves
Example 9 : Trace the curve
x =a (t + sin t), y = a(l cos t) Cycloid
Solution: The given curves are x = a(t + sin t), y= a (1  cos t)
so dx = a(l+ cos t) and dy = a sin t
dt dt
dy asint
:. dt = a(l+cost)
2 sin t/ 2cost/ 2
= ';:'
2cos
2
t/2
= tan t/2
63
A Textbook of Engineering Mathematics Volume  I
t=O n n n12 n

2
x =0
a( + 1)
an
)
an
y=O a 2a a 2a
dy =0
1
(X)
1
(X)
dx
we observe that there is a point (0,0) on the curve and slope of the tangent there
is 0 i.e. tangent will be x axis. Again we find that there is a point [ a( + 1 ),a ] and
slope being equal to 1 showing that tangent at that point will make an angle of
45
0
with positive direction of x axis. Further there is a point (an, 2a) and slope
being (X) indicates that tangent makes an angle of 900 with x axis. Similarly other
points are observed. If we go on giving to t values greater than n or less than n,
we shall have the same type of branches.
C
t =n:
A
2a
n:
t= 
2
t = n:
M an: o t=O an: L
The point 0 is called vertex and X'OX i.e. LM is tangent at the vertex, CAB is
called the base.
Example 10 : Trace the curve
x = a (t  sin t), y = a(l  cos t)
Solution:
dx ( ) dy .
 = a 1  cos t ,  = a SIn t
dt dt
64
Cllrve Tracing
dy asint 2sint/2cost/2
:. dx = a(lcost) = 2sin
2
t/2
= cott/2
t=O n/2 n 3n 2n

2
x=O
a(%l)
an
a(3; +1)
2an
y=O a 2a a 0
dy
=00
1 0 1 00
dx
Here we have taken t from 0 to 2n and we get one complete cycloid. If we give
negative values we shall get the corresponding cycloid on the other side of y axis
and so on.
t = 1t
1t
B
t= 
2
2a
a1t
XI
0 t= 0 A
X
Example 11 : Trace the curve
x = a cos t + ~ log tan
2
.!. ,y = a sin t (Tractrix)
2 2
65
A Textbook of Engineering Mathematics Volume  I
Solution:
dy . all 2 t
=asmt+2'
t
.sec 
dt 2 tan 2 2
2
. a
=asmt+.,.
2
. t t
smcos
2 2
= ~ ( 1  sin
2
t)
smt
cos
2
t
=a
sin t
dx
and = acost
dt
dy dy /dt
:. dx = dx/dt
sin t
= acost. 2
acos t
= tan t
t=O
x=oo
y=O
dy =0
dx
1t

2
0
a
00
1t 1t

2
00 0
0 a
0 00
When t =0 the corresponding point is ( 00, 0) i.e. a point at infinity on the
negative side of x axis and slope there, being 0 indicates x axis will be tangent at
that point which is situated at 00 hence x axis is an asymptotes. Similarly plot
other points.
66
Curve Tracing
EXERCISE
1. Trace the curve y2 (2a x) =x
3
(Cissoid)
Ans.
y
x=2a
o ,(2a,0)
yl
2. Trace the curve xy2 = a
2
(a  x)
Ans.
,
67
A Textbook of Engineering Mathematics Volume  I
y
x=a
XI
                  ~                 ~ ~ ~ ~ ~ X
o
yl
3. Trace the curve y2(a  x) = x2(a + x) (Strophoid)
(U .P.P.C.S. 1994)
Ans.
y
x+a=O x=a
(a, 0)
~   ~   ~ ~                 ~       + X
yl
4. Trace the curve x2y2 = a
2
(y2  X2)
Ans.
68
Curve Tracing
y
x =a
I
I
I
I
I
I
I
I
I
I
I
I
: x=a
o
Y'
5. Trace the curve X2/3+y2/3 = a
2
/
3
Ans.
(Astroid)
Y
B (0, a)
t=n t=O
X X'
(a, 0) A' A (a, 0)
3n
t =  B' (0, a)
2
Y'
6. Trace the curve r = a sin28
Ans.
69
(V.P.T.V.2oo2)
A Textbook of Engineering Mathematics Volume  I
9 = 31t
, 4
,
,
,
Y
.
,
1t
9= 
," 4
x '         ~ ~ ~ ~         ~ x
,
Y'
,
,
,
Remark: In the curve r = a sin nO or r = a cos nO, 3 n loops if n is odd and 2n if n
is even.
7. Trace the curve 3ay2 =x(x  a)2
Ans.
Y
o ~  ~       ~ X
Y'
8. Trace the curve r = a sin 39
Ans.
70
(U.P.T.U.2001)
Curve Tracing
0= 27t
0= 51t
3
6
7t
9= 
2
0= ~
3
7t
0= 
6
o = 7t        ; ~ ~ ~ = = =      0 = 0
9 = 77t
6
0= 47t
3
0= 37t
2
9. Trace the curve r = a+b cosO, a >b
(Pascal's Limacon)
Ans.
0= I07t
6
       ~           4     ~ X
o O=7t
10. Trace the curve reciprocal spiral rO =a
Ans.
71
A Textbook of Enzineering Mathematics Volume  I
Y
,
y=a
,
,
,
,
,
\
,
\
\
,
,
X' X
Y'
11. Trace the equiangular spiral r = a e
flCota
Ans.

12. Trace the curve r = 2 a case (circle)
Ans.
72
,
I
I
I
8 = 7t I
2:
Curve Tracing
o
8=0
~               ~       ~ X
2a A
2a
13. Trace the curve  = 1 + cos e (Parabola)
r
Ans.
Y
(4a, 23
1t
)
(2a, ~ )
C
(4a 2:)
3 ' 3
X'            r     +                ~ X
o A (a, 0)
Y'
Objective problems
Four alternative answers are given for each question, only one of them is correct.
Tick mark the correct answer 
1. Which of the following lines is a line of symmetry of the curve x
3
+ y3 = 3(xy2
+yx2)?
(i) x =0
(iii) Y =x
(ii) Y =0
(iv) Y = x
(LA.S.1993)
Ans. (iii)
73
A Textbook of Engineering Mathematics Volume  I
2. For the curve y2 (2a  x) = x
3
, which of the following is false?
(i) The curve is symmetrical about x axis.
(ii) The origin is a cusp.
(iii) x = 2a is a asymptote.
(iv) As x ~ 3 a , y ~ 00.
3. The curve X
2
(X
2
+ y2) =a
2
(x
2
_ y2)
(1) Symmetrical about x =0, y =0
(2) Y = x are tangents at origin
(3) x = a are the asymptotes.
(4) has no point of inflexion of these statements
(i) Only 1 and 2 are correct.
(ii) 1, 2 and 4 are correct.
(iii) 2, 3 and 4 are correct.
(iv) All are correct.
4. The equations of the inverted cycloid is 
(i) x = a(S  cosS)
y = a(1  cosS)
(ii) x = a(S + sinS)
y = a(1  cosS)
(iii) x = a(S sinS)
y = a(1  cosS)
(iv) x = a(S  sinS)
y = a(1 + cosS)
Ans. (iv)
Ans. (ii)
(v.P.P.C.S. 1994)
Ans. (iii)
5. The curve given by the equations x = a cos
3
S, y = a sin
3
S is symmetric about
(V.P.P.C.S.1995)
(i) Both the axes.
(ii) x  axis only.
(iii) yaxis only.
(iv) None of the two axes.
6. The curve x
3
+y3 3axy = 0 has at origin 
(i) Node (ii) Cusp of first species
(iii) Cusp of second species (iv) Conjugate point
Ans. (i)
Ans. (i)
(R.A.S.1994)
7. The curve y = 3x
5
 30x
4
+ 3x  20 has how many points of inflexion?
74
Curve Tracing
(i) None
(iii) Two
(ii) One
(iv) Four
8. The graph of the function f(x) =! is
x
Y
(i)
X ' ~       ~         ~ X
Y'
Y
(iii) X' ~     I      ~ X
(
Y'
Ans. (ii)
(M.P.P.CS.1995)
Y
(ii) X' ~     I      ~ X
~
Y'
Y
(iv) X' X
Y'
Ans. (ii)
(R.A.S. 1995)
9. Consider the following statements with regard to the curve x
6
+ y6 = a
2
x2y2
Assertion (A)  Curve is symmetrical about both the axes.
Reason (R)  By putting x and y in place of x and y respectively, the equation of
the curve remains unaltered.
of these statements 
(i) Both A and R are true and R is a correct explanation of A.
(ii) Both A and R are true and R is not a correct explanation of A.
(iii) A is true but R is false.
(iv) A is false but R is true.
75
A Textbook of Engineering Mathematics Volume  I
Ans. (ii)
10. The graph of the curve x = a( y = a sin t is symmetrical
about the line.
(i) x =0
(iii) Y = x
(ii) Y =0
(iv) None of these.
le 2t.
11. The graph of x = 2' Y = 2 IS a
l+t l+t
(i) Circle (ii) Ellipse
(iii) Cycloid (iv) None of these.
Ans. (ii)
Ans. (i)
12. If x and y both are the odd functions of t then the curve is symmetrical 
(i) About x axis (ii) About y axis
(iii) About y = x (iv) In opposite quadrant
Ans. (iv)
13. The curve x = a cos<l> , y = a sin<l> is
(i) Circle (ii) Ellipse
(iii) Parabola (iv) None of these
Ans. (i)
14. The curve traced by the equation y2(a + x) = x
2
(a x) is
76
Curve Tracing
Y
(i)
I
(a, 0) I (a, 0)
X'   '    ' +   ~    " " f  '    '   + X
Y'
Y
Y'
(iv) None of these.
Ans. (ii)
15. The curve traced below is represented by 
Y
(0,2a)
2a
X ' +  ~      ~      + X
o
Y'
77
A Textbook of Engineering Mathematics Volume  I
(i) X(y2 + 4a
2
) = 8a
3
(iii) y(X2 + 4a
2
) = 8a
3
(ii) xy2 = 4a
2
(2a x)
(iv) None of these
16. The curve traced by r = a(1 + cose) is
(i)
(2a, n)
(ii)
(iii)
8=n
8=0
(iv) None of these.
3at 3ae .
17. The curve traced by x = 3' Y = 2 IS
1+t 1+t
(0, n)
Y
,
,
Ans. (iii)
(2a,O)
Ans. (ii)
:x=a
,
,
(i) X' (
(a, 0)
(ii) X'+,'I*""+ X
: (a, 0)
Y'
Y
(iii)
Y'
,
Y'
Y
"
(iv) ... X
78
..............
...............
x + y + a = tf'
Y'
Ans. (iv)
Curoe Tracinz
18. The curve traced by x = a(t +sin t) and y = a(l + cos t) is
Y
2a :
(0,2a)
(i)
    ~             ~           ~     x
(a1[,O) (a1[,O)
Y'
Y
(a1[,2a) (il1[,2i1)
(ti)
X' ''====t==' X
o
Y'
Y
(iv)
X'               +                      ~ x
Y'
(iv) None of these
79
A Textbook of Engineering Mathematics Volume  I
19. The Folium of Descartes is given by the equation
(i) (x
2
+y2)2 = a
2
(x2 _ y2)
(ii) x
3
+ y3 = a
3
(iii) x
3
+y3 = 3axy
(iv) None of these
20. The Lemniscate of Bernoulli is represented by the equation
(i) x
6
+y6 = 3x2y2
(ii) (x
2
+y2)2 = a
2
(x2 _ y2)
(iii) x
5
+ y5 = 5a
2
x2y2
(iv) None of these
21. The number of leaves in the curve r = a sin 59 are
(i) Two (ii) Five
(iii) Ten (iv) None of these.
22. Match the list I with list II 
List I
curves
(a) x2y2 = x
2
 a
2
(b) y(X2 + 4a
2
) = 8a
3
(c) y2x = a
2
(x  a)
(d) x2y2 = a2(x2 + y2)
List II
Portion of the curve
(1) Lies between y =0 and y = 2a.
(2) Lies outside the lines x = a and inside the lies y = 1.
(3) Does not lie between the lines x = a, y = a.
(4) Does not lie between the lines x = 0 and x =a.
Correct match is
abc d
(i) 2 4 1 3
(ii) 2 1 4 3
(iii) 2 1 3 4
(iv) 1 2 4 3
23. Match the list I with list II.
List I
curves
(a) y2(a
2
+ X2) = x2(a
2
 X2)
80
Ans. (i)
Ans. (iii)
Ans. (ii)
Ans. (ii)
Ans. (ii)
(b) a
4
y2 = a
2
X4  x
6
(c) x
2
y2 = (a + y)2 (a
2
 y2)
(d) x
3
 ay2 =0
(e) x
3
+ y3  3axy =0
List II.
Symmetry about
(1) y = x
(2) x =0
(3)x =0, Y =0
(4) Y =0
Then correct match is
a b
(i) 3 1
(ii) 3 3
(iii) 3 3
(iv) 3 2
c
2
2
2
3
24. Match the list I with list II.
List I.
(a) r = a(1 + cos8)
(b) r = a(1 + sin8)
(c) r2 = a
2
cos 28
(d) r = a cos28
List II
Symmetrical about
(1) 8 =0 and 8 = nl2
(2) 8 =0
(3) 8 = ~
2
The correct match is
a b
(i) 2 3
(ii) 2 1
(iii) 2 3
(iv) 3 2
c
1
3
3
1
Curve Tracing
d e
4 3
1 4
4 1
4 1
Ans. (iii)
d
1
1
1
2
Ans. (i)
81
"This page is Intentionally Left Blank"
Chapter 4
Expansion of Function of One
Variable
Taylor's Theorem: If f(x +h) is a function of h which can be expanded in
ascending powers of h and is differentiable term by term any number of times
w.r.t.h then
f(x+h) = f(x) +hf'(X) + ~ fll(X)+ ~ flll(X) + ................. ~ fn(x) + .. .
Proof: Let f(x+h) = AD + A1h+ A2h2+ A3h3 + A
4
h
4
+ ................... (i)
Differentiating it successively w.r.t.h, we get
f'(X+h) = Al + 2A2h + 3A3h2 + 4A4h
3
+ ................. (ii)
fll(X+h) = 2A2 + 3.2A3h + 4.3A4h
2
+ ...................... (iii)
f'Il(X+h) = 3.2 A3 +4.3.2 A4h + ............................ (iv)
Putting h =0 in (i), (ii), (iii), (iv), ............ etc. we get
1 1 1
AD = f(x), Al = l! f'(X), A2 = ~ fll(X). A3 = ~ fill (x) etc.
Substituting these values of AD, Al, A2, .......... etc. in (i) we get
h
2
h
3
f(x+h) = f(x) + hl'(x) + ~ fll(X)+ ~ flll(X)+ ....................... (v)
Corollary 1. Substituting x =a in (v), we get
h
2
h
3
f(a +h) =f(a) + hf'(a) + ~ fll(a) + ~ f'1I(a) + ................ (vi)
Corollary 2. Substituting a =0 and h= x in (vi), we get
2 3
f(x) = f(O) + Xf'(O) + ~ fll(O) + ~ flll(O) + ........ .
Which is Maclaurin's theorem,
Corollary 3. Substituting h = x a in (vi), we get
(x_a)2 (x_a)3
f(x) = f(a) + (x a) f'(a) + ~ f"(a)+ ~ fill (a) + ......... .
This is the expansion of f(x) in powers of (x  a).
x x
2
x
3
Example 1: Show that log (x +h) = log h + h  2h2 + 3h
3

83
A Textbook Mathematics Volume  I
Solution: Here we are to expand in powers of x. Thus we have to use the form
x
2
x
3
f(x+h) = f(h) + xf'(h) + lZ f"(h) + f"'(h) + ......... (i)
Here f(x+h) = log (x +h)
112
:.f(h) = log h; f'(h) = ; f"(h) = 2 ; f'''(h) = 3 etc.
h h h
:. Substituting these values in (i) we get
log (x+h) = log h + x ( :2 )+ (:3 )+ ................ .
x x
2
x
3
= log h + h  2h2 + 3h3 + ................ .
Hence Proved
Example 2: Express the polynomial2x
3
 7x
2
+ xI in powers of (x2)
Solution:
Here f(x) = 2x
3
+ 7x
2
+ xI
:. f'(x) = 6x
2
+14x +1; f"(x) = 12 x +14 f"'(x) = 12; fiv (x) =0
Here we are to use the following Taylor's theorem
(x_a)2 (x_a)3
f(x) = f(a) + (xa)f'(a) + lZ f"(a) + f"'(a) + ................ (i)
Here a = 2, as we are to expand in powers of x2
:. putting x =2 in f(x), f'(x), ............ etc, we get
f(2) = 2(2)3 + 7(2)2 +2 1 = 45
f'(2) = 6(2)2 + 14(2) +1 = 53
f"(2) = (12) (2) +14 = 38
f"'(2) = 12, fiv(2) =0 etc.
:. From (i) substituting a =2 and these values, we get
(X_2)2 (X_2)2
2x
3
+ 7x
2
+ xI = 45 +(x 2) + .. lZ (38) + (12)
= 45 + 53 (x2) + 19(x2)2 + 2(x2)3 Answer.
Example 3: Expand sin x in powers of ( x  ) .
Solution: Here f(x) = sin x
= sin { + ( x  ) }
Here we should use the following form of Taylor's theorem.
84
(B.P.S.c. 2007)
Expansion ofFunction of One Variable
(x  a)2 (x _ a)3
f(x) = f(a) + (x a) fl(a) + fll (a) + flll(a) + ............ (i)
Now we have f(x) = sin x and a = we are to expand in powers Of( x 
fl(X) = cos x, fll(X) =  sin x, flll(X) =  cos X fiv (x)= sin x etc.
. 1t f(1t) 1t 1
putting x = '2 ' we get '2 = sm'2 =
e(i)=cosi=o; f
ll
(i)=
sin
i=1
= = Oflv = sin = 1 etc.
2 2' 2 2
From (i), putting a = and substituting these values, we get
2
. 1t
smx=1+( x'2}O)+ (1)+ (0)+ 11 (1)+ ................ .
= 1 (x  J + (x  J ..................... Answer.
EXPANSION OF FUNCTION OF SEVERAL VARIABLES:
TAYLOR'S SERIES OF TWO VARIABLES:
If f(x,y) and all its partial derivatives upto the nth order are finite and continuous
for all point (x,y) where
a::;; x::;; a+h, b::;; y::;; b+k
Then f(a+h, b+k) = f(a,b) + + + f + + f + .........
ax ay t ax ay ax ay
Proof: Suppose that f(x + h, y + k) is a function of one variable only, say x where
y is assumed as constant. Expanding by Taylorls theorem for one variable, we
have
a
f(x + Ox, Y +oy) = f(x,y +oy) + ox f(x,y + oy) +J.12 f(x,y + oy) + ............ ..
ax ax
Now expanding for y, we get
a (oy)2 a
2
f(x + Ox, Y +oy) = [f(x,y) + Oy ay f(x,y) + 'ay2 f(x,y)+ ...................... ]
85
A Textbook Mathematics Volume  I
+Ox . .i.[f(x,y)+Oyaa f(X,y)+ ........... ] + .......... ]+ ..
ax y ax ay
[
a ( ay )2 a
2
]
= f(x,y)+oy ay f(x,y)+ ay2 f(x,y)+ ............ +
[
a a
2
] (ay)2 [ a
2
]
Ox ax f(x,y)+oy axay f(x,y) + ax
2
f(x,y)+ ........... + ......... .
= f(x, y) + [ox.i.f(x, y) + oy aa f(x, y)] + [(Ox)2 a
2
f(X; y) +
ax y ax
2oxoy a
2
f(x,y) + (oyf a
2
f(X;y)]+ ........ .
axay ay
[
af af] 1 [a
2
f a
2
f a
2
f]
h
a
+k +,,) h2_2 +2hk+k2_2 + ......... .
x ax ax axay ay
=>f(a + It. b+ k) = f(a,b)+ +k +k J f+ ........... .
on putting a = 0, b =0, h= x, k =y, we get f(x,y) =
(
af af) 1 ( 2 a
2
f a
2
y 2 a
2
f)
f(O,O) + x ax + y ay + x ax2 + 2xy axay + y ay2 + ................. .
Example 1: Expand eX sin y in powers of x and y as for terms of the third degree,
Solution: Here f(x,y) = ex sin y f(O,O) =
fx(x,y) = eX sin y fx(O,O) =
fy(x,y) = eX cos y => fy(O,O) = 1
fxx(x,y) = ex sin y fxx(O,O) = 0
fxy(x,y) = eX cos y fxy(O,O) =1
fyy(x,y) =  eX sin y fyy(O,O) = 0
fxxx(x,y) = eX sin y fxxx(O,O) = 0
fxxy(x,y) = eX cos y fxxy(O,O) = 1
fxyy(x,y) =  ex sin y fxyy(O,O) = 0
fyyy(x,y) =  ex cos y fyyy(O,O) = 1
Then by Taylor's theorem, we have f(x,y)= f(O,O) +
a a 1 a a 1 a a
( ) ( )
2 ( )2
x ax + y ay f (0,0) + x ax + y ay f (0,0) + x ax + y ay f (0,0) + ........... .
86
Expansion of Function of One Variable
x
2
2xy y2 1 3
= f(O,O) + xfx(O,O) + y(O,O) + "jryfxx (0,0) + ",fxv(O,O)+ "j":;"" fvv (0,0) + x fxxx (0,0) +

3x
2
y 3 2 1 3
fxxV(O,O) + 1'2 xy fxyV(O,O) + 1'2 y fyvV(O,O) + ................. .
 .
X2 y2
sin y = + x(O) + y(l) +  (0) + xy (1) +  (0)
2 2
x
3
3x
2
y 3xy2 y3
+(0)+(1)+(0)+(1) + .............. .
6 6 6 6
. x2y y3
+ xy +26+ .............. Answer
Example 2: Expand ex cos y near the point ( 1,%) by Taylor's theorem.
(V.P.T.V 2007)
Solution: We have
( ) ( )
2 ( )3
a a 1 a a 1 a i)
f(x+h,y+k)=f(x,y)+ h+k f+ h+k f+;.; h+k f+ ........ .
ax ay ax ay ax ()y
Again ex cos y = f(x,y) = f[ 1 + (x l),%+(y  %)]
1t
where h = x1 and k= y
4
1t
ex cos y = f(l+h, '4 +k)
:. f(x,y) = ex cos y (1,%) = 52
Of( 1
of = eX cos y , 4 =
ox ox J2
of x' 4 e
ay = e sm y ay =  J2
o2f x 02f( e
ox2 = e cos y ox2 = J2
87
A Textbook of Engineering Mathematics Volume  I
in(l
a
2
f = _ex cos y => '4
a
2
f x a
2
f( e
axay = e smy => axay = 
Substituting these values in Taylor's Theorem, we obtain
eX cosy = h +[(Xl) h +(y  h)]
Jz+2(Xl)(yH Jzl]+ ..
Answer.
y
Example 3: Expand tan
1
; in the neighbourhood of (1,1) by Taylor's theorem.
Hence compute f(l.l, 0.9).
Solution: Here f(x,y) = tan
1
.................. (i)
a = 1, b = 1
:. f(l)) = tan
1
n) = tan
1
putting x = y =1 in (i)
By Taylor's theorem, we have
(U.P.T.U. 2002, 2005)
[
af af] 1 [a
2
f a
2
f a
2
f]
f(x,y)=f(a,b)+ (xa)ax +(yb)ay +ll (xa)2
ax2
+
2
(xa)(yb)axay +(yb)2
ay2
+ ......
Here a= 1, b = 1 so we have
f(x,y) = f(l))+ +(y_l)2 .... (ii)
Now from (i) we have
y 1t
f(x,y) = tan
1
; => f(l,l) = '4
:. aaxf = 1 (y ) _ y => (af) _ 1
7  x' +y' ax l'.'l 2:
a
2
f 2xy ( a
2
f ) 1
ax
2
= (X2 +y2)2 => ax
2
(1,1) ='2
88
Expansion of Function of One Variable
Substituting the values of f(l,l) etc in (ii) we have
Clx (1,1) Cly (1,1) ClxCly (1,1)
tan
1
r = l)+!(y 1) + _1)2 .!+ 2(x l)(y 1).0 +(y _1)2(_!)]+ .........
x 42 2 2 2
Y 1t (XI) (yl) 1 [(X_l)2 _(Y_l)2] ...
tanl; = 4  2 + 2 + 2 + .............. (111)
putting (xI) = 1.1 1 = 0.1, (yl) = 0.9 1 =  0.1, we get
1t 1 1 1 1
f(1.1,0.9) =4'2 (0.1) '2(0.1)+4' (0.1)2=4' (0.1)2
= 0.8862  0.1
= 0.7862 Answer.
Example 4: Expand Xy in powers of (xI) and (yl) upto the third degree terms.
(U.P.T.U 2003)
Solution: We have f(x,y) = xY
Here taking a =1 and b =1 we have Taylor's expansion as
[
Clf Clf] 1 [ 2 Cl
2
f Cl
2
f 2 Cl
2
f]
f(x,y)=f(a,b)+ (xa)+{yb) +1') (xa) 2 +2(xa){yb)+{yb) 2
Clx Cly Clx ClxCly Cly
+ (xafJ +3(xa)2{yb) +3(xa){ya)2 +(y_b)3_
1 [Cl
3
f Cl
3
f Cl
3
f Cl
3
f]
Clx' ClxClxCly ClxClyCly Cl
y
3
Now f (x,y) = Xy =>f(l,l) =1
fx (x,y) = YXyl =>fx (1,1) =1
fy (x,y) = xylog x =>fy (1,1) = 0
fxx (x,y) = y(yl)xy2 =>fxx (1,1) = 0
fxy (x,y) = xyl + yxy1log x =>fxy (0,0) = 1
fyy (x,y) = xY(log x)2 =>fyy (1,1) = 0
fxxx (x,y) = y(y1 )(y2)xy 3 =>fxxx (1,1) = 0
fxxy (x,y) = (yl)xy2 +y(yl) x y
2
1o
g
x+yxY 2 =>fxxy (1,1) = 1
89
(1)
A Textbook Mathematics Volume  I
2logx
fxyy (x,y) = YXY1 (log x)2 + xY = yxY1(log x)2+2XY1 log x =>fxyy (1,1) =
x
fyyy (x,y) = xy (log xp =>fyyy (0,0) =
Substituting these values in the Taylor's expansion, we get
1 1
xy = 1 +(xl) +0 + 11 [0+2(x1) (y 1) + 0] + [0 + 3(x 1)2 (y 1) + 0+0]
1
=>xy = 1+ (xl) + (x 1) (y 1)+ 2" (x1)2 (yl) Answer.
Example 5: Find the first six terms of the expansion of the function eX log (1 +y) in
a Taylor's series in the neighourhood of the point (0,0)
Solution: Here f(x,y) = ex log (1 +y) => f(O,O) =Q
fx(x,y) = eX log (1 +y) => fx (0,0) =0
eX
fy(x,y) =  => fy (0,0) =1
l+y
fxx(x,y) = ex log (1 +y) => fxx (0,0) =0
eX
fxy(x,y) =  => fxy (0,0) =1
l+y
_ex
fyy(x,y) = 2 => fyy (0,0) = 1
(l+y)
fxxx(x,y) = eX log (1 +y) => fxxx (0,0) =0
eX
fxxy(x,y) = => fxxy (0,0) = 1
l+y
eX
fxyy(x,y) =  2 => fxyy (0,0) = 1
(1 + y)
2e
x
fyyy(x,y) = 3 => fyyy (0,0) = 2
(1 + y)
we know that
1
f (x,y) = f(O,O)+ [x fx (0,0) + Y fy (0,0)]+ 2" [X2 fxx (0,0,)+ 2xy fxy (0,0)+ y2 fyy (0,0)] +
[x
3
fxxx (0,0) + 3x
2
y fxxy (0,0) + 3xy2 fxyy (0,0) + y3 fyyy (.0,0)] + ............ .
1 1
:. ex log (l+y) = +[x.O + y.l] + [X2. +2xy.1 + y2 (1)] +  [x
3
. + 3x
2
y.l + 3xy2
2 6
(1) + y3.2]+ .............. .
1 1 1 1
eX log (1 + y) = Y + xy   y2 +  x2y   xy2 +  y3+ ........... Answer.
2 2 2 3
90
Expansion or Function orOne Variable
EXERCISE
1. Apply Taylor's theorem to find the expansion of log sin(x + h).
h
2
h
3
Ans. log sin (x + h) = log sin x + h cot x   cosec
2
x +  cosec
2
cot x + ........ .
2 3
2. Expand 2x
3
+ 7x
2
+ x 1 in powers of (x 2).
(B.PS.C. 2007)
Ans. 45 + 53(x2) + 19(x2)2 + 2(x3)2
3. Expand x2y + 3y 2 in powers of (xI) and (y + 2) using Taylor's theorem.
(U.P.P.C.S. 1992; P.T.U. 2006)
Ans. 10  4(xl) + 4(y +2) 2(x 1)2 + 2(xl) + (y + 2) + (x 1)2 (y + 2)
(x+h)(y+k)
4. Expand ( ) ( ) in powers of hand k upto and inclusive of the 2nd
x+h + y+k
degree terms.
xy hy2 + kx
2
h 2y2 2hkxy _
Ans. + 2 + 3 3 + ................ .
x+y (x+y) (x+y)' (x+y) (x+y)
5. If f(x,y) = tan
1
(xy), compute an approximate value of f(0.9, 1.2)
Ans.  0.823
6. Find the expansion for cos x cos y in powers of x, y up to fourth order terms.
_ x2 y2 X4 x2y2 y4
Ans. cos x cos y  1  2  2 + 24 + 4 + 24 + .............. .
7. Expand sin (xy) about the point ( 1,%) upto and including second degree terms
using Taylor's series.
Ans. I_1t2 (x_l)2
8 2 2 2 2
91
"This page is Intentionally Left Blank"
UNIT 2
Differential Calculus  II
"This page is Intentionally Left Blank"
JACOBIANS
Chapter 5
Jacobians
Jacobians : If u and v are functions of the two independent variables x and y,
then the determinant.
OU ou

ox oy
ov ov
ox oy
is called the Jacobian of u, v with respect to x,y and is written as
o(u,v) J( ) J (u'V)
( )
or U,V or 
o x,y x,y
(U.P.P.C.S.1990)
Similarly if u, v and w be the functions of three independent variables x,y and z,
then
ou ou ou

ox oy oz
o(u,v,w) ov ov ov
=  
o(x,y,z) ox oy oz
ow ow ow

ox ay oz
. . J( ) J(U'v,w]
It is wntten u,v,w or
x,y,z
Properties of J aco bians
(1) First Property:
If U and v are functions of x and y, then
o(u,v) o(x,y) =1
or JJ' =1
o(x,y)'o(u,v)
Where J= o(u,v) and J'= o(x,y)
o(x,y) d(U,V)
(U.P.T.U.2005)
95
A Textbook of Engineering Mathematics Volume  I
Proof: Let u= u(x,y) and v = v(x,y), so that u and v are functionus of x and y
au au
ax ay
 
Now a(u,v) a(x,y) =
ax ay

au av
a(x,y),a(u,v)
av av ay ay

ax ay au av
on interchanging the rows and column of second determinant
au au ax ay
ax ay au au
av av ax ay
ax ay av av
au ax au ay au ax au ay
+ +
ax au ay au axav ayav
(i)
=
av ax av ay av ax av ay
+ +
ax au ay au axav ayav
Now differentiating u = u(x,y) and v =v(x,y) partially with respect to u and v, we
get
au = 1 = au . ax + au . ay
au ax au ay au
au = 0 = au . ax + au . ay
av ax av dy av
dv = 1 = dv . dX + dv . dy
dV ax dv ay dv
dv =o=dV.aX+dV.ay
du ax du dy au
(ii)
on making substitutions from (ii) in (i) we get
d(U,V) a(x,y) =11 01=1
a(x, y)' d(U, v) 0 1
or JI' = 1 Proved.
(2) Second Property
If u, v are the functions of r,s where r,s are functions of x,y then
d(U,V) _ d(U,v) d(r,s)
a(x,y)  a(r,s) d(X,y)
96
lacobians
au au
ar ar

Proof: a ( u, v) . a ( r , s) =
 
ax ay
ar as
a(r,s) a(x,y) av av as as

ar as ax ay
on interchanging the columns and rows in second determinant
au au ar as
ar as ax ax
=
av av ar as
ar as ay ay
au ar au as au ar au as
+ +
arax asax Oray csay
av ar av as
.+.
ar ax as ax
au au
ax ay
=
avav
axay
_ a(u,v)
 a(x,y)
av Or av as
+
aray csay
Proved
Note  Similarly we can prove that
a(u,v,w) _ a(u,v,w) a(r,s,t)
a(x,y,z)  a(r,s,t) a(x,y,z)
(3) Third Property
If functions u,v,w of three independent variables x,y,z are not independent, then
a(u,v,w) = 0
a(x,y,z)
Proof: As u, v, ware independent, then f (u, v, w) =0 (i)
Differentiating (i) w.r.t x,y, z we get
~ . a u + ~ . a v + ~ . a w =0 (ii)
au ax ay ax aw ax
~ . au + ~ . av + af . aw =0 (iii)
au ay av ay away
~ . a u + ~ . a v + af .aw =0 (iv)
au az av az aw az
97
A Textbook of Engineering Mathematics Volume  I
EI
" . af af af f (") ("') d (.) h
Immating ,, rom 11, 11l an IV, we ave
au av aw
au av aw
ax ax ax
au av aw=o
ay ay ay
au av aw
az az az
on interchanging rows and columns, we get
au au du
dX ay dZ
av dv dv
=0 ~  
ax ay az
dW dW ow
ax ay oz
o(u, V, w)
Proved
~ =0
a(x,y,z)
x X X
3
X
1
X x
Example 1: If Y1 =..2..l., Y2 = , Y3 = _1_2 , Show that the Jacobian of YI, Y2,
Xl X
2
X3
y3 with respect to Xl, X2, X3 is 4.
(V.P.P.C.S. 1991, V.P.T.V. 2002, 2004)
Solution: Here given
X
2
X
3
X
3
X
1
X
1
X
2
Y1 =, Y2 =, Y3=
Xl
X
2
X3
ay1 ay1 ay1
aX
1
OX
2
OX
3
O(YJlY2'Y3) =
ay2 ay2 ay2
O(X
Jl
X
2
,X
3
) aX
1
aX
2
aX
3
ay3
Y3
ay3
oX
1
aX
2
aX
3
98
=
~
x
2
~
X3
2 2 2 1 1 1
= Xl X2 X3 1 _ 1 1
X 2X 2X 2
I 2 3 1 11
= 1 (11) 1(11) + 1(1+1)
= 0 + 2+ 2
= 4 Proved
[acobians
Example 2: If X = r sinS cos<t>, y = r sinS sin<t>, z = r eosS, show that
o(x,y,z) 2' S d fi d o(r,S,<t
';'; = r sln an n
o(r,S,<t o(x,y,z)
Solution:
ox ox ox
  
or oS o<t>
o(x,y,z) ay ay
oy
=
 
o(r,S,<t or oS o<t>
oz oz dZ
 
or oS
dct>
99
(U.P.T.U.2001)
(U.P.T.U.2001)
A Textbook of Ellgineering Mathematics Volume  I
sinScosq, rcosScosq,  rsinSsinq,
= sinSsinq, rcosSsinq, rsinScosq,
cos S  r sin S 0
sin Scos q, cosScosq,  sin S
= r2 sinS sinSsinq, cosSsinq, cosq,
cosS  sinS 0
= r sm S cos S + sm S
2. { IcosScosq, sinq,1 . ISinscosq, sinq,ll
cosSsinq, cosq, sinSsinq, cosq,
= r2 sinS [cosS (cosS cos
2
q, + cosS sin2q,)+ sinS (sinS cos
2
q, + sinS sin2q,)]
= r2 sinS (cos
2
S + sin2S)
= r2 sinS
a(r,S,q,) 1 ..
:. ( ) = 2. (usmg fIrst property)
a x,y,z r smS
Answer.
a(x,y,z)
Example 3: If u = x y z, v = x
2
+ y2 + Z2, w = X + Y + z, find J = a ( ) .
U,v,W
(U.P.T.U. 2002, 03)
S I
S' l' . l' f' 1 a(u,v,w)
o utton: mce u, v, ware exp !CIt y gIVen, so Irst we eva uate J' = ( ) .
au au au
ax ay az
a a av yz zx xy
Now J' = 2 v  = 2x 2y 2z
ax ay az
111
aw aw aw

ax ay az
= yz (2y  2z)  zx (2x  2z) + xy (2x  2y)
= 2[yz (y  z)  zx (x  z) + xy (x  y)]
=2(X2y  x
2
z  xy2 + xz2 + y2z  yz2)
= 2[X2 (y  z) x (y2  Z2) + yz (y  z)]
= 2(y  z) [ x
2
 X (y + z) + yz]
= 2(y  z) [y(z  x)  x (z  x)J
= 2(y  z) (z x) (y  x)
= 2(x  y) (y  z) (z  x)
Hence by JJ' = 1, we have
a(x,y,z) 1
J= a(u,v,w) = 2(xy)(yz)(zx)
a x,y,z
Answer.
100
Tacobians
o(x,y,z)
Example 4: If u = x + y + z, uv = y + z, u v w = z, Evaluate ( )
o U,V,w
Solution:
x = u  uv = u (1  v)
Y = uv  uvw = uv (1  w)
z=uvw
ox ox ox
 
ou ov ow
. o(x,y,z) _ oy oy
~
.. o(u,v,w) 
ou ov ow
oz oz oz
ou ov ow
Iv u
=
v(1 w) u(l w)
vw uw
1 0
=
v(1 w) u(l w)
vw uw
Applying R1 ~ R 1 + (R2 + R3)
= u
2
v (1  w) + u
2
vw
= u
2
v Answer.
(LA.S. 2005, U.P.P.CS. 1990, U.P.T.U. 2003)
0
uv
uv
0
uw
uv
Example 5: If u, v, ware the roots of the equation (I..  xp + (I..  yp + (I..  zp =0, in
... f' d o(u,u,w)
fl., In ( ) .
o x,y,z
(LA.S. 2002, 2004; U.P.P.CS. 1999; B.PS.C 2007; u.P.T.U. 2002)
Solution: Given
(I..  xp + (I..  y)3 + (I..  z)3 =0
~ 31..
3
 3(x + Y + Z)1..2 + 3(x
2
+ y2 + Z2)1..  (X3 + y3 + Z3) =0
sum of the roots = u + v + W = x + y + z (i)
product of the roots = uv + vw + wu = x
2
+ y2 + Z2 (ii)
1 (1 3 3) (1'1'1')
uvw =  x + y' + z
3
Equation (i), (ii) and (iii) can be written as
f1 = u + v + ro  x y  Z
f2 = UV + vw + wu  x
2
 y2  Z2
101
A Textbook of Enzineerinz Mathematics Volume  I
1
fJ = uvw  (x
3
+ y3 + x
3
)
3
af
l
af
l
af
l
ax ay az
a( fl ,f2/ f3) af
2
af
2
af
2
a(x/y/z) = a; dy az
af
3
af
3
af
3
ax ay az
1 1 1
= 2x 2y 2z
_x
2
_y2 _Z2
1 1 1
= ( ~ 1 ) (  2 ) (  1 ) x Y z
x
2
y2 Z2
o o
=2 xy yz
x
2
_ y2 y2 _ Z2
o
=2(xy)(yz) 1
o
1
1
1
z
x+ y y +z Z2
= 2 (x  y) (y  z) (y + Z  x  y)
= 2 (x  y) (y  z) (z  x)
af
l
afl af
l
au av aw
a(fl/f2/f3) af
2
af
2
af
2
a(u/v/w) = au av aw
af
1
af
3
af
1
au av aw
Applying
102
[acabians
1 1 1
=
v+w u+w u+v
vw wu uv
0 0 1
=
vu wu u+v
w(v  u) u(wv) uv
0 0 1
=(vu)(wv) 1 1 u+v
w u uv
=(v  u) (w  v) (u  w)
=  (u  v) (v  w) (w  u)
0(fl'f2'(')
Th f
o(u,v,w) o(x,y,z)
ere ore =
o(x,y,z)
a(u,v,w)
2(x  y)(y z)(z  x)
=
(u  v)(v w)(w  u)
2(xy)(yz)(zx)
= Answer.
(u  v)((v  w)(w  u))
EXERCISE
_ _. o(x,y) o(r,e)
1. If x  r cose, y  r sme, evaluate ( ) and ()
o r,e a x,y
_ u+v. o(u,v)
2. Ifx  uv, y = ,fmd ( )
uv ox,y
o(u,v) 1 (y2 _X2)
3. If u
3
+ v
3
= X + y, u
2
+ v
2
= x
3
+ y3, prove that =
o(x,y) 2 uv(uv)
1
Ans. r, 
r
A
(u  V)2
ns. '''
4uv
4. If u = xy + yz + zx, v = x
2
+ y2 + Z2 and w = x + y + z, determine whether there
is a functional relationship between u,v,w and if so, find it
103
A Textbook of Engineering Mathematics Volume  I
a(u,v,w)
Ans. w
2
 v  2u = 0, = 0
a(x,y,z)
5. If u,v,w are the roots of the equation in A and _x_ + _y_ + _z_ = 1
a+A b+A C+A
a(x,y,z)
Then find a ( )
u,v,w
6. If U,V,W are the roots of the equation
a(u, v, w)
(x  ap + (x  bp + (x  cp =0 then find ~    :  '
a(a, b,e)
104
(LA.S. 2005)
A
(uv)(vw)(wu)
ns.  '   :   ~  ~     ,  : . . .
(ab)(bc)(ca)
(UP.T.U.2002)
Ans _ 2(a b)(bc)(ca)
. (uv)(vw)(wu)
Chapter 6
Approximation of Errors
Let z = f(x,y) (i)
If ox, oy are small increments in x and y respectively and oz, the corresponding
increment in z, then
z + oz = f(x + ox, y + oy) (ii)
Subtracting (i) from (ii), we get
oz = f(x + Ox, Y + oy)  f(x,y)
of of
= f(x,y) + ox  + oy  + .............. f(x,y)
ox ay
af af .
=ox+oy ::l" (approxImately)
ax vy
As neglecting higher powers of Ox, oy
af af .
:. oz = ox + oy (approxImately)
ax oy
~ If ox and oy are small changes (or errors) in x and y respectively, then an
approximate change (or error) in z is oz.
Replacing Ox, oy, oz by dx, dy, dz respectively, we have dz = ~ dx + ~ dy
ax ay
Note 1. If Ox is the error in x, then relative error = ox
x
Ox
percentage error =  x100
x
Note 2. f'(x) dx is called the differential of f(x)
Example 1 The time T of a complete oscillation of a simple pendulum of length L
is governed by the equation T = 2nJ,
Find the maximum error in T due to possible errors upto 1 % in Land 2% in g.
(U.P.T.U. 2004,2009)
Solution: We have T = 2nJ
1 1
log T= log 2n + logeL  log .. g
2 2
105
A Textbook ot :::ngineering Mathematics Volume  I
Differentiating
dT =O+.:!. dL _.:!. dg
T 2 L 2 g
=O( :g )XIOO 1
But dL
x100
=1 dg xl00=2
L ' g
so
dT 1 3
T x 100 = 2"[
1
2] = 2"
Maximum error in T = 1.5% Answer.
Example 2: The power dissipated in a resistor is given by P = E2 . Find by using
R
calculus the approximate percentage change in P when E is increased by 3% and
R is decreased by 2 %.
E2
Solution: Here given P = R
Taking logarithm we have
log P = 2 log E  log R
on differentiating, we get
oP oR
PER
or 100 oP = 2 x 1000E _ 1000R
PER
oP
or lOOp = 2(3)  (2)
=8
Percentage change in P = 8% Answer.
Example 3: In estimating the number of bricks in a pile which is measured to be
(Sm x 10m x Sm), count of bricks is taken as 100 bricks per m
3
. Find the error in
the cost when the tape is stretched 2% beyond its standard length. The cost of
bricks is Rs. 2,000 per thousand bricks.
Solution: We have volume V = xyz
Taking log of both sides, we have
log V= log x + log Y + log z
Differentiating, we get
106
(V.P.T.V. 2000, 2004)
Approximation of Errors
oV ox oy oz
=++
V x Y z
100 oV = 100 Ox + 100 oy + 100 oz
V x y z
= 2 + 2 + 2 (As given)
=6
=> oV = 6 (5x10x5)
100 100
= 15 cubic metre
Number of bricks in oV = 15
x
l00
= 1500
E
. t 1500 x 2000
rror In cos
1000
=3000
This error in cost, a loss to the seller of bricks = Rs. 3000. Answer.
Example 4: What error in the common logarithm of a number will be produced
by an error of 1 % in the number.
Solution: Consider x as any number and, let
y = loglo X
1
Then oy = loglo e ox
x
ox
=log10 e
x
= (0: x 100)(
1
= IOglOe
100
ox
:.as given  xl00 = 1
x
0.43429
=
100
= 0.0043429
which is the required error. Answer
Example 5: A balloon is in the form of right circular cylinder of radius l.5m and
length 4m and is surmounted by hemispherical ends. If the radius is increased by
0.01 m and length by 0.05m, find the percentage change in the volume of balloon.
Solution: Here given,
radius of the cylinder (r) = l.5m
length of the cylinder (h) = 4m
or = O.Olm, oh = 0.05m
(U.P.T.U. 2002, 2005)
107
A Textbook of Mathematics Vall/me  I
Let V be the volume of the balloon then.
V= volume of the circular cylinder + 2 (volume of the hemisphere)
= nr
2
h + 2 ( j nr
3
)
4
= nr
2
h +  nr
3
3
I
I
I
14m
I
I
I
I
I
I
4
Now BV = n2rBrh, h + nr2 Bh + n.3r2 Br
3
= nr[2hBr + rBh + 4rBr]
BV nr[2(h+2r)Br+rBhJ
 = !='_':_=
V nr
2
h + inr3
3
2(h + 2r )Br + rBh
= '__ ':c__
4 2
rh+r
3
2( 4 + 3)(0.01) + (1.5)(0.05)
4 0
(1.5x4)+:3 (l.5t
0.14 + 0.075 0.215
= =
6+3 9
:. BV x 100 = 0.215 x 100 = 21.5 = 2.389% Answer.
V 9 9
Example 6: If the base radius and height of a cone are measured as 4 and 8 inches
with a possible error of 0.04 and 0.08 inches respectively, calculate the percentage
(%) error in calculating volume of the cone.
Solution:
1
Volume V= nr
2
h
3
Taking log,
[U.P.T.U. (C.O.) 2003]
108
Approximation of Errors
1
log V = log  + log n + 2 log r + log h
3
Differentiating, we get
BV = 2 Or + oh
V r h
=> 0: =
= 0.03
:. Percentage (%) error in volume
= 0.03 x 100
=3% Answer.
Example 7: Find the percentage error in the area of an ellipse when an error of +1
percent is made in measuring the major and minor axes.
Solution: If x and yare semimajor and semiminor axes of ellipse. Then its area
A is given by
A = nxy
Taking log,
log A = log n + log x + log y
Differentiating, we get
oA = 0+ Ox + oy
A x y
or OAX100=oxx100+0Y.XI00
A x y
= 1+1
=2
:. Error in the area = 2 % Answer.
Example 8: Find the possible percentage error in computing the parallel
resistance r of three resistance rl, r2, r3 from the formula.
1 1 1 1
=++
rr,r
2
r
3
If rl, r2, r3 are each in error by + 1.2 %
Solution: we have
1 1 1 1
=++ (i)
r r
1
r
2
r3
Differentiating, we get
(U.P.T.U.1999)
109
A Textbook of Engineering Mathematics Voll/me  I
1 1 1 1
eSr=eSr eSr eSr
2 212223
r r
1
r
2
r3
=>!(eSr X100)+.!(
eSr
3 X100)
r r r
1
r
l
r
2
r
2
r3 r3
= .!(1.2) + .!(1.2) + .!(1.2)
r
1
r
2
r3
= 1.2('! +.! +.!)
r
1
r
2
r3
= 1.2(;) from (i)
Or
=> x100 =1.2
r
Hence % error in r = 1.2 Answer.
Example 9: The angles of a triangle are calculated from the sides a,b,c. If small
changes eSa, eSb and eSc area made in the sides, find eSA, eSB and eSc approximately,
where is the area of the triangle and A, B, C are angles opposite to sides a, b, c
respectively. Also show that eSA + eSB + eSc = 0
Solution: We know that
a
2
= b
2
+ c
2
2abc cos A (i)
Differentiating eqn (i), we get
2aeSa = 2beSb + 2coc  2b& cos A  2 ceSb cosA + 2bc sin A eS A
or be sin A eSA = aeSa (b  c cos A) eSb  (c  b cos A) eSc
(U.P.T.U.2002)
or = aeSa  (a cos C + C cos A  c cos A) eSb  (a cos B + b cos A  b cos A) &
a
or eSA=  (eSa  eSb cos C  eSc cos B) (ii)
similarly, also by symmetry, we have
b
oB =  (eSb  eSc cos A  eSa cos C) (iii)
c
and eSc =  (eSc  eSa cos B  eSb cos A) (iv)
Adding equations (ii), (iii) and (iv) we get
1
eSA + eSB + eSc =  [(a  b cos C  c cos B) eSa + (b  c cos A  a cos C) eSb + (c  a cos
B  b cos A)eSc]
[(a  a) eSa + (b  b) eSb + (c  c) eSc]
110
Approximation of Errors
"1
= (0+0+0)
2A
=0
Thus, oA + oB + oC =0 Hen,'e Proved.
Example 10: Two sides a, b of a triangle and included angle C are measured,
show that the error oe in the computed length of third side c due to a small error
in the angle C is given by
oC (a sin B) = &
Solution: we know that
a
2
+ b
2
+ c
2
cosC = ,where a, b are fixed while c and C vary
2ab
. Cl:C 2coc
:.Sln u =
2ab
=> oe = ab sinCoC (i)
c
Again by sine formula, we have
b c
=
sinB sinC
. B bsinC
=>Sln =
c
Proved.
Example 11: If A be the area of a triangle, prove that the error in A resulting from
a small error in e is by
oA = A [.!. + _1_ + _1 ___ 1_Joc (U.P.T.U. e.O. 200607)
4 s sa sb sc
Solution: we knIJW that
A = .j:,(sa)(s b)(se)
1
log A ='2 [log s + log (s  a) + log (s  b) +log (s  c)]
Differentiating w.r.t.c, we get
.!. oA = .!.[.!. os + 1 o(s  a) +_1_0(S  b) +_1_0(S  c)]
A & 2 s oe (sa) & (sb) oc (se) oc
1
Now s =  (a + b +e)
2
os 1
111
(i)
A Textbook of Engineering Mathematics Volume  I
1
Also s  c =  (a + b  c)
2
. o(sc) 1
.. oc =2
1 o(sa) 1
similarly, s  a =  (b + c  a) =:) =
2 oc 2
1 o(sa) 1
andsb=  =:) =
2 oc 2
From (i)
1[1(1) 1 (1) 1 (1) 1 ( 1)]
; 2 +sa 2 +sb 2 +(sc)2
___ I_]oc
4 s sa sb sc
Proved.
Example 12: If the kinetic energy T is given by T = !mv
2
find approximate the
2
change in T as the mass m change from 49 to 49.5 and the velocity v changes from
1600 to 1590.
Solution: Here given
T=!mv
2
2
1
:. oT = 2 {(om) v
2
+ m (2vov)}
or oT = ! v
2
om + mv ov (i)
2
It is given that m changes from 49 to 49.5
:. om = 0.5 (ii)
Also, v changes from 1600 to 1590
:. ov = 10 (iii)
Hence, from (i)
oT =! (1600)2 (0.5) + 49(1600) (10)
2
= 144000
Thus, T decreases by 144000 units. Answer.
Example 13: Find approximate value of [(0.98)2+ (2.01)2 + (1.94)2]1/2
Solution: Let f(x, y, z) = (X2 + y2 + z2)1/2 (i)
Taking x =1, y =2 and z = 2 so that dx =  0.02, dy = 0.01 and dx =  0.06 from (i)
af af af
ax = x (X2 + y2 + Z2) 1/2, ay = Y (X2 + y2 + Z2)1/2, az = z (X2 + y2 + Z2) 1/2
112
Approximation of Errors
Now df = af dx + + dz (by total differentiation)
ax ay uZ
= (X2 + y2 + Z2)1/2 (xdx + ydy + zdz)
1
=  (0.02 + 0.02  0.12)
3
=  0.04
:. [(0.98)2 + (2.01)2 + (1.94)2]1/2 = f (1,2,2) + df
= 3 + (0.04)
= 2.96 Answer.
Example 14: In determining the specific gravity by the formula S = where
Aw
A is the weight in air, w is the weight in water, A can be read within 0.01 gm and
w within 0.02 gm. Find approximately the maximum error in S if the readings are
A = 1.1 gm, w = 0.6 gm. Find also the maximum relative error.
Solution: Given
Aw
Differentiating above, we have
(Aw)BAA(OABw)
uS= 2
(Aw)
(i)
Maximum error in S can be obtained if we take oA =  0.01 and ow= 0.02
:. From (i)
(1.1 0.6)(0.01) (1.1)(0.010.02)
 (1.10.6)2
== 0.112
Maximum relative error in
S = (BS)max = 0.112
S (1.1)
1.10.6
= 0.05091 Answer.
EXERCISE
1. The deflection at the centre of a rod, of length 1 and diameter d supported at its
ends and loaded at the centre with a weight w varies as w1
3
d4. What is the
percentage increase in the deflection corresponding to the percentage increases in
w, I and d of 3%,2% and 1 % respectively.
113
A Textbook of Engineering Mathematics Volume  I
Ans.5%
2. The work that must be done to propel a ship of displacement 0 for a distance 5
.,. . I 5
2
0
2
/
3
P' d . I th . f k
In time t IS proportiona to 2' In approXImate y e Increase 0 wor
t
necessary when the displacement is increased by 1 %, the time diminished by 1 %
and the distance diminished by 2 %.
Ans. !%
3
3. The indicated horse power of an engine is calculated from the formula
PLAN nd
2
I = 32000 ,where A;:: 4 . Assuming that errors of r percent may have been
made in measuring P, L, Nand d, find the greatest possible error in 1 %.
Ans.5r%.
4. In estimating the cost of a pile of bricks measured as 6m x 50m x 4m, the tape
is stretched 1 % beyond the standard length. If the count is 12 bricks in 1m
3
and
bricks cost Rs. 100 per 1,000 find the approximate error in the cost.
(U.P.T.U.2005)
Ans. Rs 43.20
5. If the sides and angles of a triangle ABC vary in such a way that ~ t s circum
d
. t P th oa ob oc 0
ra IUS remaIns constan, rove at + + ;::
cosA cos B cosC
6. Compute an approximate value of (1.04)301
Ans.1.12
7. Evaluate [(4.85)2 + 2(2.5)3)1/
5
Ans.2.15289
8. The focal length of a mirror is given by the formula .!  .! = ~ . If equal errors 0
v u f
are made in the determination of u and v, show that the relative error Of in the
f
focal length is given by B ( ~ + ~ ) .
114
Chapter 7
Extrema of Functions of Several
Variables
Introduction: There are many practical situations in which it is necessary or
useful to know the largest and smallest values of a function of two variables. For
example, if we consider the plot of a function f(x, y) of two variables to look like a
mountain range, then the mountain tops, or the high points in their immediate
vicinity, are called local maxima of f(x, y) and the valley bottom, or the low points
in their immediate vicinity, are called local minima of f(x, y). The highest
mountain and deepest valley in the entire mountain range are known as the
absolute maxima and the absolute minimum respectively.
Absolute maximum
z
Local maximum
y
Absolute minimum
Definitions
Definition (i): A function f of two variables has a local maximum at (a, b) if f(x, y)
:5; f(a, b) when (x, y) is in neighbourhood of (a, b). The number f(a, b) is called
local maximum value of f.
115
A Textbook of Engineerillg Mathematics Volume  I
If f(x, y) f(a, b) for all points (x, y) in the domain of the function f, then the
function has its absolute maximum at (a, b) and f(a, b) is the absolute maximum
value of f.
Definition (ii)
If f(x, y) f(a, b) when (x, y) is near (a, b) then f(a, b) is the local minimum value
of f. If f(x, y) f(a, b) for all points (x, y) in the domain of f then f has its absolute
minimum value at (a, b).
Theorem: If f has a local maximum or minimum at (a, b) and the first order
partial derivatives of f exist at a, b, then fx(a, b) =0 and fy (a, b) =0
Definition (iii): A point (a, b) in the domain of a function f(x, y) is called a critical
point (or stationary point) of f(x, y) if fx (a, b) =0 and fy (a, b) =0 or if one or both
partial derivatives do not exist at (a, b).
Definition (iv): A point (a, b) where f(x, y) has neither a maximum nor a
minimum is called a saddle point to f(x, y).
Necessary Condition for Extremum Values of functions of two Variables:
Let the sign of f(a + h, b + k)  f (a, b) remain of the same for all values (positive or
negative) of h, k. Then we have
(i) For f (a + h, b + k)  f (a, b) <0, f (a, b) is maximum
(ii) For f (a + h, b + k)  f (a, b) >0, f (a, b) is minimum
By Taylor's theorem for two variables, we have
f(a + h, b + k) = f(a, b) +
(
af af) 1 ( 2 a
2
f af 2 a
2
f)
h; + k; + I" h ;T + 2hk :;; + k ;T + ...... ,
uX oy (.,b) t.f uX oxuy oy (o,b)
or f(a+h, b+k)  f(a, b) = (h af) + + ................ terms of second and
. ax (o,b) ay (o,b)
higher orders in hand k (1)
By taking hand k to be sufficiently small, we can neglect the second and higher
order terms. Thus, the first degree terms in hand k can be made to govern the
sign of the left hand side of equation (1). Therefore the sign of [f (a+h, b+k)  f(a,
1 (2)
VJ
Taking k =0, we find that if :;:. 0 the right hand side of equation (2)
ax (o,b)
changes sign whenever h change sign. Therefore, f(x, y) cannot have a maximum
or minimum at x =a, y = b if :;:. 0
oX (o,b)
116
Extrema of Functions of Several Variables
similarly, taking h =0, we can find that f(x, y) cannot have a maximum or a
minimum at x =a, y = b if (:f ) :;c 0
y (d.b)
Thus, the necessary condition for f(x, y) to have a maximum or a minimum at x =
a, y = b is
(
af ) = 0 and ( ~ ) = 0
ax (d.b) i)y (d.b)
This condition is necessary but not sufficient for existence of maxima or minima.
Now, from equation (1)
1 [i)2f i)2f i)2f]
f(a+h, b+k)  f(a, b) = l h
2
ax2 + 2hk i)xdy + k
2
i)y2
1
f(a+h, b+k) f(a, b) = II [h2r + 2hks + k
2
t] (3)
i)2f i)2f
s=and t=2 at (a, b)
i)xi)y i)y
since, we assumed that the sign of the LHS of equation (3) is the sign of the RHS
of the equation (3) i.e.
sign of LHS of equation (3) = sign of [rh
2
+ 2hks + k
2
t]
= sign of .!. [r2h2 + 2hkrs + k2rt]
r
= sign of .!. [(r2h2 + 2hkrs + k
2
s2) + ( k
2
s
2
+ k2rt)]
r
= sign of ~ [ ( h r + kS)2 + k
2
(rt  S2) ]
= sign of .!. [k
2
(rt  S2)]
r
.: (hr + ks)2 is always positive
= sing of r if rt  S2 >0
Hence, if rt  S2 >0, then f(x, y) has a maximum or a minimum at (a, b) according
to r < 0 or r > 0 respectively.
Working rule to find Extremum Values:
The above proposition gives us the following rule for determining the maxima
and minima of functions of two variables.
117
A Textbook of Enzineerinz Mathematics Volume  I
Find ~ and af Le'. p and q equate then to zero. Solve these simultaneous
ax cry
equations for x and y Let aI, bI; a2, b2; , ........ be the pairs of roots.
. a
2
f a
2
f a
2
f
Fmd ax
2
' axay' oy2' (Le. r, t and s respectively) and substitute in them by
turns aI, bl; a2, b2; ......... for x, y. Calculate the value of rt  S2 for each pair of
roots.
If rt  r
2
> 0 and r is negative for a pair of roots, f(x,y) is a maximum for this pair.
If rt  r2 > 0 and r is positive, f(x, y) is a minimum. If rt  S2 < 0, the function has a
saddle point there.
If rt  S2 =0, the case is undecided, and further investigation is necessary to decide
it.
Example 1: Find the local maxima or local minima of the function f(x, y) = x
2
+ y2
 2x  6y + 14
Solution: We have
f(x,y) = x
2
+ y2  2x  6y + 14
Therefore
fx(x, y) = 2x 2 and fy(x, y) = 2y  6
If we put fx(x, y) = 0 we have
2x 2 =0
=>x=l
If we put fy(x,y) =0, we have
2y  6 =0
=>y=3
we get x =1 and y = 3 as critical points of function f(x, y). Now, we can write
f(x, y) as
f(x, y) = 4 + (x 1)2 + (y  3)2
since (x 1)2 ~ 0 and (y  3)2 ~ 0, we have f(x, y) ~ 4 for all values of x and,
therefore, f(l, 3) =4 is a local minimum. It is also the absolute minimum of f.
Example 2: Find the extreme value of f(x,y) = y2  x
2
Solution: we have
f(x, y) = y2  x
2
Differentiating partially with respect to x and y, we get
fx = 2x, fy = 2y
which follows fx(O,O) =0 and fy (0,0)=0
Therefore, (0,0) is the only critical point.
However the function f has neither local maximum nor a local miruma at (0,0).
Thus (0,0) is called saddle point of f.
Example 3: Find the maximum and minimum of the function f(x,y) = x
3
+ y3_
3axy
118
Extrema of Functions of Several Variables
(V.P.T.V. 2004, 2006; u.p.p.es. 1992; B.p.s.e 1997)
Solution: Given that
f(x,y) = x
3
+ y3  3 axy (i)
Therefore
df df
p = dX = 3x
2
 3ay, q = dy = 3yL 3ax
~ f ~ f ~ f
r = dX2 = 6x, s = dXdy = 3a, t = oy2 = 6y
for maxima and minima, we have
af
 = =>3x
2
 3ay =0 => X2 = ay
ax
df
 = =>3y2  3ax =0 => y2 = ax
dy
(iii)
putting the value of y in equation (iii), we get
X4 = a
3
x
=>x (X3  a
3
) =0
=>x (x  a) (x
2
+ ax + a
2
) =0
=>x = 0, x = a
(ii)
Putting x =0 in (ii), we get y =0, and putting x =a in (ii) we get y =a. Therefore
(0,0) and (a, a) are the stationary points (i.e. critical points) testing at (0, 0).
r =0, t = 0, s = 3a => rt S2 = Negative
Hence there is no extemum value at (0,0).
Testing at (a, a)
r = 6a, t =6a, s = 3a
=> rt  S2 = 6a x 6a  (3a)2
= 36a
2
 9a
2
= 27a
2
> and also r = 6a >
Therefore (a, a) is a minimum point.
The minimum value of f(a, a) = a
3
+ a
3
 3a
3
=  a
3
Answer.
Example 4: Test the function f(x, y) = x
3
y2 (6  x  y) for maxima and minima for
points not at the origin.
Solution:
Here f(x,y) = x
3
y2 (6  x  y)
= 6 x3 y2 _ x4y2 _ X3
y
3
df
:. p = ax = 18x
2
y2  4 x3y2  3x2y3
119
A Textbook of Engineering Mathematics Volume  I
of
q =  = 12 x
3
Y  2x4y  3X3y2
Or
02f
r =  = 36 xy2 12 x2y2  6xy3
ox
2
= 6xy2 (6  2x y)
02f
S = = 36 x2y  8x3y  9x2y2
oxoy
= x2y(36  8x  9y)
02f
t =  = 12 x
3
 2X4  6x
3
y
oy2
= x
3
(12 2x  6y)
Now, for maxima or minima, we have
of
ax = 0 => x2y2 (18  4x  3y) =0 (i)
and of =0
Or
=> x3y(12  2x  3y) =0 (ii)
From (i) & (ii)
4x+ 3y =18
2x + 3y =12 and x =0 =y
solving, we get x = 3, Y =2 and x =0 = y. Leaving x =O=y, we get x =3, y =2. Hence
(3, 2) is the only stationary point under consideration.
Now,
rt  S2 = 6x
4
y4 (6  2x y) (12  2x  6y)  X4 y2 (36  8x  9y)2
At (3, 2)
rt  S2 = + ive (> 0)
Also, r = 6(3) (4) (6  6  4)=  ive 0)
:.f(x,y) has a maximum value at (3, 2).
Example 5: Examine for minimum and maximum values. sin x + sin y + sin (x +
y)
Solution: Here, f(x,y) = sin x + sin y + sin (x + y)
of
:. p =  = cos x + cos(x + y)
ox
of
q = Or = cos y + cos( x + y)
iff . . ( )
r =  =  sm x  sm x + y
ox
2
120
(U.P.P.C.S.1991)
Extrema of Functions of Several Variables
iff .
s = =  sm (x + y)
axay
a
2
f . . ( )
t =  =  sm y  sm x + y
ay2
Now
ax ay
=> cos x + cos (x + y) =0 ........... (i) and cos y + cos (x + y) =0 ........ (ii)
subtracting (ii) from (i) we have
cos x  cos y =0 or cos x = cos y
=>x=y
From (i), cos x + cos 2x =0
or cos 2x =  cos x = cos (1t  x)
1t
or 2x = 7t  X :. x = 
3
1t . . .
:. x = y = "3 IS a stationary pomt
At r =  fj  fj = fj s = fj
3'3 ' 2 2 ' 2
t = _ fj  fj =fj
2 2
3 9
:. rt  S2 = 3   =  > 0
4 4
Aslo r < 0
:. f(x, y) has a maximum value at
Maximum value = f(
. 1t . 1t . 21t
= Sln + SIn + SIn 
333
fj fj J3 3fj
=++=
2 2 2 2 Answer.
Example 6: Test the function f(x,y) = (X2 + y2)e<h\,2) for maxima or minima for
point not on the circle x
2
+ y2 =1
Solution:
Here f(x, y) = (X2 + y2) e<x2+y2)
121
A Textbook of Engineering Mathematics Volume  I
. af _ (x2 + y2) e(x';.y') (2x) + 2x e(X';'Y')
.. ax 
= 2x (1 x2 _ y2) e(X';.v')
af (' 2) (' ')
 = (x2 + y2) e x;.y (2y) + 2y e x ;.y
ay
= 2y (1 x2 _ y2) e(x';.y')
Now for maxima and minima, we have
af af
=0 and =0
ax ay
~ 2x (1 x2  y2) e (' +y') = 0 and 2y (IxL y2) e(x' +v') =0
~ X =0, Y =0 and x
2
+ y2 =1
Leaving x
2
+ y2 =1 as given, take x =0, y =0
Hence (0,0) is the only stationary point
fx = (2x  2x3  2xy2) e(x';.y')
a
2
f ) (x';.,,') (x'...,.')
r = ax
2
= (2  6x
2
 2y2 e . + (2x  2x
3
_2xy2) e . (2x)
= e(X';.y') (4X4 10 x2 + 4x2y2  2y2 +2)
s = fxy= ( 4xy) e(X'+Y') + (2x  2x3  2xy2) e(X'+Y') (2y)
= ( 8xy + 4 x3y + 4xy3) e(x
2
+y')
a
2
f " ' ,
t =  = e(x ;.)' ) (2  2x
2
 6y2) + e(x;.y) . (2y  2yx2  2y3) (2y)
ay2
= e(X';.y') (2 2x
2
10y2  4x2y2 + 4y4)
At (0,0), r =0, s =0, t =2
rts2 = 4 > 0
Also r = 2 (> 0)
:. f(x, y) has a minimum value at (0,0)
:. Minimum value = (0+0) e 0 =0 Answer. .
Example 7: A rectangular box, open at the top, is to have a given capacity. Find
the dimensions of the box requiring least material for its construction.
Solution: Let x, y and z be the length, breadth and height respectively, let V be
the given capacity and 5, the surface
V is given =:> V is constant
V=xyz
V
or Z= (i)
xy
5 = xy + 2xz + 2yz
122
Extrema ofFunctions of Several Variables
2V 2V f( ) , (')
=xy++= x,y usmg 1
y x
ill 2V ill 2V
:,p==y,q==x
ax x
2
ay y2
a
2
f 4V a
2
f
r== s==l
ax
2
Xl' axay
a
2
f 4V
t==
ay2 y3
af af
Now =0 and =0
ax dy
2V (") d 2V 0 ("')
~ y  2 = 0 ........ " II an x  2 = ........ " III
X Y
From (i) y = 2 ~
x
4
:, From (ii), x  2V 4 ~ 2 = 0
or x ( 1  ; ~ ) = 0
or x = (2V)1/3 (As x ::I: 0)
and = 2V = 2V = (2V)I/3
Y x
2
(2V)3/2
:, x =y = (2V)1/3 is a stationary point. At this point, r = 4V = 2 > 0, s =1,
2V
t= 4V = 2
2V
So that rt  S2 = 4 1 = 3> 0 and r > 0
~ S is minimum when x =y = (2V)1/3
V V
Also z =  = 2/3
xy (2V)
V
I
/3 (2Vt3
= 2
2
/
3
= 2
=I.
2
Hence 5 is minimum when x =y = 2z = (2V)1/3 Answer,
Example 8: Find the semivertical angle of the cone of maximum volume and of a
given slant height.
123
A Textbook of Ensineerins Matllematics Volume  I
(U.P.T.U.2005)
Solution: Let 1 be the slant height and S be the semi  vertical angle of the cone.
Then, radius of the base, r = OC = 1 sinS and height of the cone, h = OA = 1 cosS.
1
Let V be the volume of the cone then V =  nr
2
h
= .!. n(1 sinS)2 (1 cosS)
3
= .!. n 1
3
sin2S cosS
3
3
A
:. dV = .!.n l2{sin2S ( sinS) + cosS. 2 sinS cosS}
dS 3
= .!.n 1
3
sinS (2cos
2
S  sin2S)
3
d
2
'; =! n 13 [sinS ( 4 cosS sinS  2 sinS cosS) + (2 cos
2
S  sin2S) cosS]
dS 3
= .!.n 13 [ 6 sin28 cosS + cosS (2 cos
2
S  sin2S)]
3
For max and min. dV = 0
dS
Which gives either sinS =0 or tan
2
S =2
so that S = 0 or S = tan
1
J2
when S = 0, volume of cone becomes zero and the cone becomes a straight line
which is not the case, when S = tan
1
J2 , we have
d
2
V 1 3 [ 2 1 ]
dS2 = "3
nl
6'"3' F3 + 0
=  ive
Hence the volume of cone is maximum when S = tan
1
J2 .
When S =  tan
1
J2, volume of cone becomes negative which is meaningless,
hence is not the case. Answer.
124
Extrema of Functions of Several Variables
Example 9: Find the volume of largest parallelopiped that can be inscribed in the
ellipsoid
x
2
y2 Z2
++=1
a
2
b
2
c
2
(U.P.T.V.2001)
Solution: Let (x,y,z) denote the coordinates of one vertices of the parallelopied
which lies in the positive octant and V denote its volume so that
V = 8xyz As 2x, 2y and 2z be the length, breadth and height respectively
:. V2 = 64X
2y
(2
Z
2 2 y2)
= 64 x2y2 c2 1  ~  
a
2
b
2
125
A Textbook of Etzzineerinz Mathematics Volume  I
At ( ~ , ~ ), rt  S2 > 0 and r < 0
Hence f(x,y) is max at( ~ , Js)
2 2 [a
2
b
2
1 a
4
b
2
1 a
2
b
4
]
=> V
max
=64c 3'37'"9'3 b2 x3
x
"9
=
27
Answer,
126
Chapter 8
Lagrange's Method of Undetermined
Multipliers
In many problems, a function of two or more variables is to be optimized,
subjected to a restriction or constraint on the variables, here we will consider a
function of three variables to study Lagrange's method of undetermined
multiplierso
Let
u = f (x, y, z) (i)
be a function of three variables connected by the relation
eI> (x, y, z) =0 (ii)
The necessary conditions for u to have stationary values are
ou =0, au =0, au =0
ax ay az
Differentiating equation (i), we get du =0 i.eo
d
au d au d au d (1
0
1
0
1
0
)
U= x+ y+ z=o
ax ay az
Differentiating equation (ii) we get del> =0 i.eo
del> = ael> dx + ael> dy + ael> dz = 0 (iv)
ox dy oz
Multiplying equation (iv) by A and adding to equation (iii) we get
(
ou + A aeldX + (au + A aeld
Y
+ (ou + A deldZ = 0
ax ox dy dy az az
This equation will be satisfied if
au + A d4> = 0 (v)
ox ax
dU + A 04> = 0 (vi)
dy dy
ou + A d4> = 0 (vii)
dZ dZ
127
A Textbook of Engineering Mathematics Volume  I
where I.. is the Lagrange multiplier. On solving equations (ii), (v) (vi) and (vii), we
get the values of x, y, z and I.. which determine the stationary points and hence
the stationary values of f(x, y, z).
Note: (i) Lagrange's method gives only the stationary values of f(x, y, z). The
nature of stationary points cannot be determined by this method.
(ii) If there are two constraints $1(X, y, z) =0 & $2 (x, y, z) =0, then the auxiliary
function is F(x, y, z) = f(x, y, z) +I..jt $, (x, y, z) + 1..2 $2 (x, y, z) here 1..1 and 1..2 are the
two Lagrange multipliers. The stationary values are obtained by solving the five
equations Fx = 0, Fy = 0, Fz = 0, FA, = 0 and FA, = 0
Example 10: Find the volume of largest parallelopiped that can be inscribed in
2 2 2 '
the ellipsoid .; + .;. + .;. = 1 using Lagrange's method of Multipliers.
abc
x
2
y2 Z2
Solution: Let 2' + 2' + 2'" = 1
abc
x
2
y2 Z2
${x Y z)=++l=O (i)
" a
2
b
2
c
2
(LA.S. 2007; U.P.T.U. 2001,2003)
Let 2x, 2y, and 2z be the length, breadth and height, respectively of the
rectangular parallelopiped inscribed in the ellipsoid. Then
V = (2x) (2y) (2z) = Bxyz
Therefore, we have
av + "A a$ = 0 =:} Byz + "A 2: = 0 (ii)
ax ax a
av + I.. a$ = 0 =:} Bxz + "A 2x = 0 (iii)
ay ay b
2
av a$ 2z
a;+"Aaz =O=:}Bxy+"A?"=O (iv)
Multiplying (ii), (iii) and (iv) be x, y and z respectively, and adding, we get
[
X2 y2 Z2]
24 xyz + 2"A 2 + 2' + 2'" = 0
abc
24 xyz +2"A (1) =0
~ I . . = 12xyz
putting the value of I.. in (ii) we have
2x
Byz + (12 xyz) 2 = 0
a
128
3x
2
=>1=0
a
a
=>x= J3
Lagrange I S Method of Undetennined Multipliers
Similarly, on putting A =  12 xyz in equation (iii) and (iv) we get
b c
y= J3'z= J3
Hence, the volume of the largest parallelopiped = 8xyz
= 8 ( ~ ) ( ~ ) ( ~ )
8abe
= 3J3
Answer.
Example 11: Find the extreme value of x
2
+ y2 + Z2, subjected to the condition xy +
yz + zx = p.
Solution: Let f = x
2
+ y2 + Z2 and cP = xy + yx + zx po
Then for maximum or minimum, we have
df + A dCP = 0 => 2x + A (y + z) = 0 (i)
dX dX
df +A dCP =0=>2Y+A(X+Z)=0 (ii)
dy dy
~ + A dCP =0=>2Z+A(X+Y)=0 (iii)
dZ dz
(U.P.T.U.2oo8)
Multiplying equation (i) by x, equation (ii) by Y and equation (iii) by z and
adding we get.
2(X2 + y2 + Z2) + 2A. (xy + yz + zx) =0
=> 2f + 2A.p =0
f
or A=
p
on putting this value of A in (i) (ii) and (iii), we get
f
2x   (y + z) =0
P
f
2y   (x + z) =0
P
f
2z   (x + y) =0
p
129
A Textbook Mathematics Volume  I
f
or 2x  2y +  (x  y) =0
P
 y) =0
f
=2and x = y
p
Similarly, we get y = z, therefore, f = 3x
2
But
xy + yz + zx =p
=> 3x
2
= P => x
2
= p/3
Therefore, extrema occur if
X2 + y2 + Z2 = P Answer.
Example 12: Show that the rectangular solid of maximum volume that can be
inscribed in a given sphere is a cube.
(U.P.T.u. 2004)
Solution: Let 2x, 2y, 2z be the length, breadth and height of the rectangular solid
and r be the radius of the sphere
Then x
2
+ y2 + Z2 = R2 (i)
Volume V= Sxyz (ii)
Consider Lagrange1s function
F(x,y,z) = Sxyz + ').. (X2 + y2 + Z2  R2)
For stationary values,
dF=O
=> (Syz + ').. (2x)} dx + (Sxz + ').. (2y)} dy + (Sxy + ').. (2z)} dz =0
=> Syz + 2Ax =0 (iii)
Szx + 2Ay =0 (iv)
Sxy + 2Az =0 (v)
From (iii) 2A x
2
=  Sxyz
From (iv) 2')..y2 =  Sxyz
From (v) 2')..Z2 =  Sxyz
:. 2A x
2
= 2')..y2 = 2')..z2
or X2 = y2 = Z2 or x =y = z
Hence rectangular solid is a cube. Hence Proved.
EXERCISE
1. Find all relative extrema and saddle points of the function.
f(x,y) = 2x
2
+ 2xy + y2  2x  2y + 5
130
Method of Undetermined Multipliers
2. A rectangular box, open at the top, is to have a volume of 32 cc. Find the
dimensions of the box requiring the least material for its construction.
(U.P.T.U.2005)
Ans. I = 4, b = 4 and h = 2 units
3. Divide 24 into three parts such that the continued product of the first part, the
square of the second part, and the cube of the third part is maximum.
Ans. x = 12, Y = 8, z =4
4. Show that the rectangular solid of maximum volume that can be inscribed in a
given sphere is a cube.
(U.P.T.U.2003)
5. The sum of three positive number is constant. Prove that their product is
maximum when they are equal.
6. Find the maximum and minimum distances of the point (3, 4, 12) from the
sphere.
Ans. Minimum distance = 12
Maximum distance = 14
7. The temperature T at any point (x, y, z) in space is T = 400 xyz2. Find the
highest temperature at the surface of a unit sphere x
2
+ y2 + Z2 =1
Ans.50
8. A tent of given volume has a square base of side 2a and has its four sides of
height b vertical and is surmounted by a pyramid of height h. Find the values of a
and b in terms of h so that the canvas required for its construction be minimum.
x
2
y2 Z2
9. Prove that the stationary values of u = 4 + 4 + 4"
abc
J5
Ans. a=h
2
h
andb=
2
x
2
y2 Z2
where Ix + my + nz =0 and 2'" + 2 +"2 = 1 are the roots of the equation.
abc
12a
4
m
2
b
4
n
2
c
4
2 + 2 =0
lau lbu lcu
(B.P.S.C. 2007)
10. Use Lagrange's method of undetermined multipliers to find the minimum
value of x
2
+ y2 + z2 subjected to the conditions.
x + y + z =1, xyz + 1 =0
Ans.3
11. Find the minimum value of x
2
+ y2 + Z2, given that ax + by + cz =p
131
A Textbook of Engineering Mathematics Volume  I
2
Ans p
. a
2
+ b
2
+ c
2
12. Find the maximum and minimum distances from the origin to the curve.
X2 + 4xy + 6y2 = 140
[U.P.T.U. (CO) 2003]
Ans. Maximum distance = 21.6589
Minimum distance = 4.5706
13. If u = ax
2
+ by2 + cz
2
and X2 + y2 + Z2 =1 and Ix + my + nz =0 are the
constraints, Prove
12 m
2
n
2
++=0
au bu cu
OBJECfIVE PROBLEMS
Four alternative answers are given for each question, only one of them is correct,
tick mark the correct.
1. Maxima and minima occur 
(i) Simultaneously (ii) Once
(iii) Alternately (iv) rarely
2. The minimum value of I x
2
~ 5x + 21 is
(i) 5
(iii) 1
(ii) 0
(iv) 2
3. The maximum value of .1 (sin x  cos x) is
(i)1
(
"') 1
111 
J2
(ii) J2
(iv) 3
Ans. (iii)
(M.P.P.C.S.1991)
Ans. (ii)
(M.P.P.C.S.1991)
Ans. (i)
4. The triangle of maximum area inscribed in a circle of radius r is
(i) A right angled triangle with hypotenuse measuring 2r
(ii) An equilarteral triangle
(iii) An isosceles triangle of height r
(iv) Does not exist
5. Let f(x) = 1 x I , then
132
(I.A.S. 1993)
Ans. (ii)
(I.A.S. 1993)
Lagrange's Method of Undetermined Multipliers
(i) f'(O) =0
(ii) f(x) is maximum at x =0
(iii) f(x) is minimum at x =0
(iv) None of the above.
Ans. (iv)
(I.AS. 1990)
6. If Y = a log x + bx
2
+ x has its extremum value at x = 1 and x = 2 then
(i)a=2,b=1
(iii) a = 2, b =
(iv) None of these
7. The function f(x) = x
3
 6x
2
+ 24x + 4 has
(i) A maximum value of x =2
(ii) A minimum value of x =2
(iii) A maximum value at x =4 and minimum value at x =6
(iv) Neither maximum nor minimum at any point
Ans. (iii)
(I. AS. 1990)
Ans. (iv)
8. The function sin x (1 + cos x) is maximum in the interval (O,n), when 
(U.P.P.C.S. 1994, M.P.P.C.S. 1995)
(i) x = n/4 (ii) x = n/2
(
... ) n
III x =
3
(
.) 2n
IV X=
3
Ans. (iii)
9. A necessary condition for f(a) to be an extreme value of f(x) is that
(i) f(a) =0
(iii) f"(a) =0
(ii) f'(O) =0
(iv) f"'(a) =0
10. The value of function
1
f(x) = x+
x
at the points of minimum and maximum one respectively
(i) 2 and 2
(iii) 1 and 1
(ii) 2 and2
(iv) 1 andl
11. The profit function is
133
(R.AS. 1995)
Ans. (ii)
(M.P.P.C.S. 1995)
Ans. (iv)
A Textbook of Engineering Mathematics Volume  I
1
P(x) =   x
2
+ 32x  480
2
then the profit is maximum if the number of item (x) produced and sold is
(U.P.P.C.S.1991)
(i) 18
(iii) 24
(ii) 17
(iv) 32
Ans. (iv)
12. The value of x for which the function f(x) = x l/x has a maximum is given by
(R.A.S. 1993)
(i) x =e
(
") 1
11 X =
e
(iii) x =e (iv) x =!
e
13. The maximum value of ~ is
(i) e
(iii) e lie
(ii) e e
(iv) e1/e
x'
14. If x + Y = k, x > 0 , y >0, then xy is the maximum when
(i) x = ky
(iii) x = Y
(ii) kx = Y
(iv) None of these
15. Maximum value of a sine + b cose is
(i) ~ a 2 + b
2
(iii).Ja  b
(ii) .Ja + b
(i v) 2 ~ r a 2=+c b2=+a+b
16. Maximum value of sin x + cos x is
(i) 2 (ii) J2
(iii) 1 (iv) 1 + J2
Ans. (i)
(M.P.P.C.S. 1994)
Ans. (iv)
(M.P.P.C.S.1991)
Ans. (iv)
(M.P.P.C.S. 1992)
Ans. (i)
(M.P.P.C.S.1992)
Ans. (ii)
17. If the functions u, v, w of three independent variables x, y, z are not
independent, then the Jacobian of u, v, w with respect to x, y, z is always equal to
(LA.S. 1995)
134
Lawange's Method of Undetermined Multipliers
(i) 1 (ii) 0
(iii) The Jalonbian of x, y, z w.r.t u, v, w
(iv) infinity
Ans. (ii)
18. The maximum rectangle, inscribed in a circle of radius 1, is of area
(U.P.P.C.S.1995)
(i) 1 (ii) 2
(iii) 4 (iv) 8
Ans. (ii)
19. The derivative f'(x) of a function f(x) is positive or zero in (a, b) without
always being zero. Then, which of the following is true in (a, b)
(i) f(b) < f(a) (ii) f(a) < f(b)
(iii) f(b) f(a) = f'(c) (iv) f(b) = f(a)
Ans. (ii)
20. The conditions for f(x, y) to be maximum are
(i) rt  S2 > 0, r >0 (ii) rt  S2 > 0, r < 0
(iii) rt  S2 < 0, r > 0 (iv) rt  S2 < 0, r < 0
Ans. (ii)
21. The conditions for f(x, y) to be minimum are
(i) rt  S2 > 0, r < 0 (ii) rt  S2 > 0, r > 0
(iii) rt  S2 < 0, r > 0 (iv) rt  S2 < 0, r < 0
Ans. (ii)
22. The stationary points of f(x, y) given by 
(i) fx =0, fy =0 (ii) fx ;f. 0 , fy =0
(iii) fy;f. 0, f(x) = 0 (iv) None of these
Ans. (i)
135
"This page is Intentionally Left Blank"
UNIT3
Matrices
"This page is Intentionally Left Blank"
Chapter 9
Matrices
INTRODUCTION: Today, the subject of matrices is one of the most important
and powerful tool in mathematics which has found application to a large number
of disciplines such as Engineering, Economics, Statistics, Atomic Physics,
Chemistry, Biology, Sociology etc. Matrices are a powerful tool in modern
mathematics. Matrices also play an important role in computer storage devices.
The algebra and caluclus of matrices forms the basis for methods of solving
systems of linear algebraic equations, for solving systems of linear differential
equations and for analysiS solutions of systems of nonlinear differential
equations.
Determinant of a matrix:  Every square matrix A with numbers as elements has
associated with it a single unique number called the determinant of A, which is
written det A. if A is n x n, the determinant of A is indicated by displaying the
elements ali of A between vertical bars
a
OI
a
22
......... a)"
det A = .  
ani a
n2
........... a
nn
The number n is called the order of determinant A.
Non Singular and singular Matrices:
A square matrix A = [al)] is said to be non singular according as I A I 7: a or I A I =
a
Adjoint of a matrix: Let A =[aij] be a square matrix. Then the adjoint of A,
denoted by adj (A), is the matrix given by adj (A) = [A'iIxn' where AI) is the
cofactor of ail in A, i.e. if
all a
l2
.......... a
1n
a
21
a
22
......... a
2n
A=
ani a,,2 ........... a
nn
139
A Textbook of Engineering Mathematics Volume  I
[
All
A21
Then adj(A) ~ ~
111
............
AI1
I]
............ A
n2
...............
............ A
nn
Inverse of a matrix : Let A be square matrix of order n. Then the matrix B of
order n if it exists, such that
AB = BA = In
is called the inverse of A and denoted by Al we have
A(adjA) = I A I I
(adjA)
or A ~ = I ,Provided I A 1"* 0
1 adjA .
or A =W' If IAI"*O
Theorem: The inverse of a matrix is unique.
Proof: Let us consider that Band C are two inverse matrices of a given matrix,
say A
Then AB = BA = I :. B is inverse of A
and AC=CA=I
C (A B) = (CA) B
or CI = IB
or C= B
:. C is inverse of A
by associative law
Thus, the inverse of matrix is unique.
Existence of the Inverse : Theorem: A necessary and sufficient condition for a
square matrix A to possess the inverse is that I A 1"* O.
Proof: The condition is necessary
Let A be an n x n matrix and let B be the inverse of A.
Then AB = In
.. I AB I = I In I = I
:. IAIIBI=1 :. I AB I = I A I I B I
:. I A I must be different from O.
Conversely, the condition is also sufficient.
If I A 1"* 0, then let us define a matrix B by the relation
B = I ~ I (adj.A)
Then AB = A [ I ~ I adjA )
140
(LA.5. 1973)
Matrices
= I ~ I (adj.A) = 1 ~ I I A I I n = In
Similarly BA = [ I ~ I adj.A )A = I ~ I (adj.A)A
1
= IAIIAIIn = In
Thus AB = BA = In
Hence, the matrix A is invertible and B is the inverse of A
Reversal law for the inverse of a Product, Theorem: If A, B be two n rowed non
singular matrices, then AB is also non  singular and
(AB)l = B1 A1
i.e. the inverse of a product is the product of the inverse taken in the reverse
order.
(LAS. 1969, U.P.P.C.S. 1995)
Proof: Let A and B be two n rowed non singular matrices,
We have I AB I = I A I I B I
Since I A I * 0 and I B I * 0
therefore I AB 1* O. Hence the matrix AB is invertible
Let us define a matrix C by the relation C = B1A1
Then C (AB) = (B1 A1) (A B) = B1 (A1A) B
= B1 In B
= B1 B = In
Also (AB) C = (AB) (B1 A1) = A (BB1) Al
= A In Al = AAl
= In
Thus C (AB) = (AB) C = In
Hence C = B1 Al is the inverse of AB.
Elementary Row Operations and Elementary Matrices
When we solve a system of linear algebric equations by elimination of unknown,
we routinely perform three kinds of operations: Interchange of equations,
Q1ultiplication of an equation by a nonzero constant and addition of a constant
multiple of one equation to another equation.
When we write a homogeneous system in matrix form AX = 0, row k of A lists
the coefficients in equation K of the system. The three operations on equations
correspond respectively, to the interchange of two rows of A, multiplication of a
row A by a constant and addition of a scalar multiple of one row of A to another
row of A We will focus on these row operations in anticipation of using them to
solve the system.
141
A Textbook of Engineering Mathematics Volume  I
DEFINITION: Let A be an nXn matrix. The three elementary row operations that
can be performed on A are
1. Type I operation: interchanging two rows of A.
2. Type II operation: multiply a row of A by a non zero constant.
3. Type III operation: Add a scalar multiple of one row to another row.
The rows of A are m  vectors. In a type II operation, multiply a row by a non
zero constant by multiplying this row vector by the number. That is, multiply
each element of the row by that number. Similarly in a type III operation, we add
a scalar multiple of one row vector to another row vector.
Inverse of Non  Singular Matrices Using Elementary Transformations:
If A is non singular matrix of order n and is reduced to the unit matrix In by a
sequence of E  row transformations only, then the same sequence of E  row
transformations applied to the unit matrix In gives the inverse of A (i.e.Al).
Let A be a non singular matrix of order n. It is reduced to unit matrix In by a finite
number of E row transformations only. Here, each E  row transformation of the
matrix A is equivalent to pre  multiplications by the corresponding E  matrix.
Therefore, there exist elementary matrices say, El, E2, ................ Er such that
[Er, E
r
l, ..................... E2, El] A = In
Postmultiplying both sides by Al, we obtain
[Er, E
r
l, ..................... E2, El] A Al = In Al
~ [Er, E
r

l
, ..................... E
2
, El] In = Al
.: AAl = In
In Al = Al
or Al = [Er, E
r

1
, ..................... E2, E1] In
Working rule to find the inverse of a non  singular matrix:
Suppose A is a non singular matrix of order n, then first we write
A= InA
Next, we apply E row transformations to a matrix A and InA till matrix A is
reduced to In. Then B is equal to Al, i.e.
B = Al
Example 1 : Find by elementary row transformation the inverse of the matrix
l ~ : ~ l
(U.P.T.U. 2000,2003)
Solution: The given matrix is
A = l ~ : ~ l
142
we can write the given matrix as
A=IA
0] [15/2
o = 4
1 5/2
~ ] = [ 1 ~ !
1 5/2
i.e. I = BA where B = Al
11/2
3
3/2
1/2
3
3/2
Matrices
or A
1
= 4 3
[
1/2 1/2
1 Answer
1/2]
5/2 3/2 1/2
Example 2 : Find by elementary row transformations the inverse of the matrix
[
~ = ~ :]
o 1 1
(U.P.T.U.2002)
143
A Textbook of Engineering Mathematics Volume  I
Solution: The given matrix is
we can write the given matrix as
A= IA
i.e.
o 1 1 0 0 1
Applying elementary transformations, we get
0
0] [1 1 0]
3 4 = 0 1 0 A, R I 7 R1  R2
1 1 0 0 1
0
0] [1 1 0]
3 4 = 2 3 0 A, R2 7 R2  2R1
1 1 0 0 1
0
][ 1 1 0]
1  4 / 3 = 2 / 3 1 0 A, R2 / 3
1 1 0 0 1
0
] [1 1 01
1  4 / 3 = 2 / 3 1 0 A, R3 7 R3 + R2
0 1/3 2/3 1 1
0
][ 1 1 01
1  4 / 3 = 2/3 1 0 A, R3 7 R3 X ( 3)
1 1 2 33
0
1
R, > R,
1 3
0 1 2 3 3
1= BA
Where B = Al
A 1 = ;  1 Ans.
2 3 3
144
Matrices
Normal Form:
Every non  zero matrix of order m x n with rank r can be reduced by a sequence
of elementary transformations to any of the following forms
1. Ir
2. [Ir,O]
3. [ ~ ]
4. [ ~ ~ ]
The above forms are called normal form of A r so obtained is a number called the
rank of matrix A
Equivalence of matrix:
Suppose matrix B of order mxn is obtained from matrix A (of the same order as
B) by finite number of elementary transformations on A; then A is called
equivalent to B i.e AB. Matrices A and B have same rank and can be expressed
as B = PAQ, where P and Q are non singular matrices. If A is of order mxn, then
P has order mXm and has nxn such that
B == PAQ
Working rule: Let A be a matrix of order mXn
1. we write A = 1m A In
2. Next, we transform matrix A to normal form using elementary
transformations.
3. Elementary row transformation is applied simultaneously to A and 1m i.e.
the prefactor matrix.
4. Elementary column operation applied to A is also applied to In i.e. the post
factor matrix.
5. Finally, we find B= PAQ, where B is the normal form of A
Example 3: Reduce matrix A to its normal form, where
A = [ ~ : ; :j
1 2 3 4
1 2 67
Hence, find the rank.
Solution: The given matrix is
(LAS 2006; V.P.T.U. 2001,2004)
145
A Textbook of Engineering Mathematics Volume  I
: ; :]
Applying elementary row transformation, we have
[
1 2  1 4] R2 Rz _ 2Rl
A
0 0 5  4 R3 R3 Rl
= 0 0 4 0' R4 R4 + Rl
o 0 5 3
Now, applying elementary column transformation we have
[
1 0 0 0] C2 C2 _ 2Cl
o 0 5  4 C3 C3 + Cl
A  0 0 4 0' C4 C
4
 4Cl
o 0 5 3
Interchanging C2 and C3, we have
A  !  C3 C2
o 5 03
o 5 0 4 R3 R3   Rz
[
1 0 0 0 ] 4
A  is' R. +
n
146
Matrices
0 0
1
1 0
5
0 1 5
0 0
16 R3
A
00 1
which is the required normal form.
Here, we have three non zero rows. Thus the rank of matrix A is 3. Ans.
Example 4 : Find non  singular matrices P,Q so that PAQ is a normal form,
where
1 3
i1
3 1
1 1
(U.P.T.U. 2002)
Solution: The order of A is 3 x 4
Total number of rows in A = 3, therefore consider unit matrix b.
Total number of columns in A == 4, hence, consider unit matrix 14
:. A3 x 4 == h A 14
0 0
1 3
61 [1
0
1 0
3 1 2 = 0 1
0 1
1 1 2 0 0
0 0
0 0
1 1
0
1 0
3 1 1
0 1
1 3 6 1 0
0 0
0 0
1 1
21[0
0
1 0
R2 R2 3R
1
6 2 4  0 1
0 1
R3 R3  2Rl
1 5 10 1 0
0 0
147
A Textbook of Engineering Mathematics Volume  I
i
: i
i
i]
As by C3 C3  5C2
& C4 C
4
 10C2
I
I 0
o 1
o 0
N=PAQ
o
o
1
o
o
1
1
o
o
1
28 28
o
o
1
28 28
4 8]
5 10 1
R3
1 0 28
o 1
1
2 A 0 0
9
28 0 0
9
1
2
A [00
1
 10
1
 ! 0 ]
o As by C
4
C4  2C3
1 2
28 0 0 0 1
o
P= 1
3
o
o
1
  _! 0]
,Q  0 0 1 _ Answer
14 28 28
o 0 0 1
Example 5 : Find the rank of the matrix
148
A=[:
10
16
1
2
3
4
Matrices
Solution: Sometimes to determine the rank of a matrix we need not reduce it to
its normal form. Certain rows or columns can easily be seen to be linearly
dependent on some of the others and hence they can be reduceds zeros by E 
row or column transformations. Then we try to find some nonvanishing
determinant of the highest order in the matrix, the order of which determines the
rank.
We have the matrix
1 3
8l
2 6
R3 R3  R2  Rl
0 0
0 0
Since I: =8# 0
Therefore rank (A) = 2 Answer
Alter
The determinant of order 4 formed by this matrix
[
: ! 1
 10 3 9 7
16 4 12 15
t ; 11
=0 :. R3 and are identical.
A minor of order 3
"[:
1
:H:
1
]bYR,>R,R2
2 2
10 3 1
=0
In similar way we can prove that all the minors of order 3 are zero.
149
A Textbook of Engineering Mathematics Volume  I
A minor of order 2
I: ~ I =8*0
Hence rank of the matrix = 2 Answer.
Example 6 : Prove that the points (Xl, Yl), (X2, Y2), (X3 , Y3) are collinear if and only
[
Xl Yl 1]
if the rank of the matrix x
2
Y 2 1 is less than 3.
X3 Y3 1
(U.P.PC.S.1997)
Solution: Suppose the points (Xl, Yl), (X2' Y2), (X3, Y3) are collinear and they lie on
the line whose equation is
ax + by + c = 0
Then
aXl + bYl + c = 0 (i)
aX2 + bY2 + c = 0 (ii)
aX3 + bY3 + C = 0 (iii)
Eliminating a, band c between (i), (ii) and (iii) we get
[
Xl Yl 1]
x
2
Y2 1 = 0
X3 Y3 1
Thus the rank of matrix
[
Xl
A= x
2
X3
is less than 3.
Conversely, if the rank of the matrix A is less than 3, then
[
Xl Yl 1]
x
2
Y2 1 = 0
X3 Y3 1
Therefore the area of the triangle whose vertices are (Xl, Yl), (X2, Y2), (X3, Y3) is
equal to zero. Hence three points are collinear.
Consistent system of Equations:
A non homogeneous system AX = B is said to be consistent if there exists a
solution. If there is no solution the system is inconsistent.
For a system of non  homogeneous linear equations AX = B (where A is the
coefficient matrix) and C = [A B] is an augmented matrix:
1. If r (A) = r (C), the system is inconsistent
150
Matrices
2. If r (A) = r (C) = n (number of unknowns) the system has a unique
solution.
3. If r (A) = r (C)< n, the system has an infinite number of solutions.
The above conclusions are depicted in figure as given below
A system of non homogeneous lineur equations
AX=B
Find r(A) and r(C)
r (A) = r (C) r (A) # r (C)
Solution exists system
is consistent
No, solution exist system
is inconsistent
r (A) = r (C) = n
IUnique solution I
r (A) = r (C) <n
Infinite number of
solutions
Example 7 : Using the matrix method, show that the equations 3x + 3y + 2z = 1;
x + 2y = 4 lOy + 3z = 2 2x  3y  z = 5 are consistent and hence obtain the
solution for x, y and z
Solution: The given system of linear equations can be written as
AX= Bi.e.
[ ~ i } l [ ~ ] + ~ l
The augmented matrix is
151
(U.P.T.U.2000)
A Textbook of Engineering Mathematics Volume  I
c = [A' B] =
. 0
2
3
2
10
3 1
Applying elementary row transformations to C, we have
J lJ.
3 2 '
R2 3Rl
1 3
:
o 29 / 3  116 / 3 '
o 17/3 68/3
_: 1
4
1]
o 1  4 '
o 1 4
:
o 1 4 '
000
10
R2
3
7
R2
3
Thus r(C) = r (A) = 3 hence the given system is consistant and has a unique
solution
or z =  4, 3y + 2z = 11, x + 2y = 4
152
Matrices
1
or z =  4, Y = 3" (2z + 11) = 1, x = 4  2y = 2
Thus, the solution is
x = 2, Y = 1, Z =  4 Answer.
Example 8 : Examine the consistency of the following system of equations and
solve them if they are consistent Xl + 2X2  X3 = 3; 3XI  X2 + 2X3 = 1 ; 2Xl  2X2 + 3X3
= 2 ; Xl  X2 + X3 =  1
(U.P.T.T:J.2002)
Solution: The given system of linear equatins can be written in matrix form as A
X = B i.e.
[; U:J[j]
The augmented matrix is
C = [A: B]
ie
Applying elementary row transformations to the augmented matrix, we obtain
!]
o 6 5 4
o 3 24
1
5
5/7
1/7
i
i.e. r (C) =.3 = r (A)
R23 Rl
R3 R3  2RI
R4  Rl
Hence, the system is consistent and has a unique solution, thus
153
A Textbook of Engineering Mathematics Volume  I
=> Xl + 2X2  X3 = 3,  7X2 + 5X3 =  8 and X3 = 4
=> 7 X2 = 5X3 + 8
1
=> X2 =  (5 x 4 + 8) = 4
7
and Xl = 3  2X2 + X3 = 3  8 + 4 = 1
Thus, the solution is Xl = 1 X2 = 4, X3 = 4 Answer
Example 9 : Examine the consistency of the following system of linear equations
and hence, find the solution 4 Xl  X2 = 12;  Xl + 5X2  2 X3 = 0;  2 X2 + 4X3 =  8
(U.P.T.V.2005)
Solution: The given equations can be written in matrix form as
[
4  1 1[X
1
1_[121
1 5 2 x
2
 0
o 2 4 X3 8
i.e. AX = B where
Now the augmented matrix C is
C = [A: B]
[
4 1 0
i.e.c=; _5
2
:
8
Applying elementary row transfomations to matrix C to reduce it to upper
triangular form we get
10
2
1'
o 2 4 8
[ =:}H
154
14
19
o
6
8
60/19
Matrices
12 1
12 ,
128/19
Hence, we see that ranks of A and Care 3 i.e. r (A) = 3 = r (C). The the system of
linear equations is consistent and has a unique solution. Thus, the given system
of linear equations is
[
1
o 0 60/19 X3 128/19
i.e. Xl + 14 X2  6 X3 = 12
19 X2  8X3 =12
60 128 128
and x =::
19 3 19 3 60
. 32
I.e. X3 =::  15
on putting the value of X3 in 19x2  8 X3 = 12
x
2
=:: + 8x
3
) =:: + 8(32))
19 . 19 15
4
=::
2 15
Lastly, putting X2, X3 in Xl + 14 X2  6 X3 = 12 we have
Xl + 14 x (  6( =:: 12
44
=:: 15
Therefore, the solution is
44 4 32
XI =::, x
2
=::, X3 =:: Answer
15 15 15
Example 10 : For what values of A and Il, the equations X + Y + Z = 6 ; x + 2y + 3 z
= 10; x + 2y + Az = Il have (i) no solution (ii) unique solution and (iii) infinite
solutions. (LA.S 2006, U.P.T.u. 2002)
Solution: The given system of linear equations can be written as
[:
i.e. AX =:: B
155
A Textbook of Engineering Mathematics Volume  I
1
2
2
1 61
3 10
A Il
Applying elementary row transformations to C, we get
B  l ~
1 1
1 2
~ ~ J
R2 ~ R2  RI
R 3 ~ R3RI
1 A1
 l ~
1 1
6 1
1 2 4 ,
A3
1l
1O
(i) For no solution, we must have
r (A);t: r (C)
i.e A  3 =0 or A = 3 and Il  10 ;t: 0 => Il ;t: 10
(ii) for unique solution, we must have
r (A) = r (C) = 3
i.e. A  3;t: 0 => A;t: 3
and Il  10 ;t: 0 => Il ;t: 10
(iii) for infinite solutions, we must have
r(A) = r (C) < 3
i.e. A  3 =0 => A = 3
and 1l10 =0 => Il = 10 Answer.
Solution of Homogeneous system of Linear  Equations:
A system of linear equations of the form AX = is said to be homogeneous where
A denotes the coefficient matrix and 0 denotes the null vector i.e.
all Xl + al2 X2 + ................. + al
n
Xn =0
a21 Xl + a22 X2 + ................. + a2n Xn =
am,x
l
+a
m
,x
2
+ ...................... +amnxn =0
i.e. AX = 0
[
all a12 .......... aln][XI] [0]
a
2l
a
22
.......... a
2l
X
2
0
or. . =.
. ..
. ..
am I a
m2
......... a
mn
Xn 0
The above system has m equations and n unknowns. We will apply the matrix
method to find the solution of the above system of linear equations. For the
156
Matrices
system AX = 0 , we see that X = 0 is always a solution. This solution is called
null solution or trivial solution. Thus a homogeneous system is always consistent.
We will apply the techniques, already developed for non homogenous systems of
linear equations to homogeneous linear equations.
(i) If r (A) = n (number of unknown) the system has only trivial solution.
(ii) If r (A) <n, the system has infinite number of solutions.
The figure as given below shows a flow chart which depicts the procedure for the
solution of a homogenous system of linear equations.
A system of homogeneous linear equations
AX=O
Always has a solution
Find r (A)
r (A) = n r (A) <n
Unique solution or Infinite no. of nontrivial
trivial solution solution
Example 11 : Find the solution of the following homogeneous system of linear
equations
Xl + X2 + 2X3 + 3X4 =0; 3 Xl + 4X2 + 7X3 + 10X4 =0
5Xl + 7X2 + llx3 + 17x4 =0; 6 Xl + 8X2 + 13x3 + 16x4 =0
Solution: The given system of linear equations can be written in matrix form as
[ ~ ~ f ~ ~ g H ~ l
Applying elementary row transformations, we get
157
A Textbook of Enzineering Mathematics Volume 1
i.e. r(A) = 4 = number of variables
R2 R2  3Rl
R3 R3  5 Rl
R.t R4  6Rl
Hence the given system of homogeneous linear equations has trivial solution i.e.
Xl = 0 = X2 = X3 = X4 Answer.
Linear Combination of vectors:
Let Xl, X2 ................ Xk be a set of k vectors in Rn. Then the linear combination of
these k vectors is sum of the form alXl + a2 X2 + ...................... + ak Xk, in which
a) is a real number.
Linear Dependence and Independence Vectors:
Let Xl, X2 ................ Xk be a set of k vectors in Rn. Then the set is said to be
linearly dependent if and only if one of the k vectors can be expressed as a linear
combination of the remaining k vectors.
If the given set of vectors is not linearly dependent, it is said to be set of Hearly
independent vectors.
Example 12: Examine for linear dependence (1, 0, 3, 1), (0, 1, 6, 1) and (0, 2, 1, 0)
in R4.
Solution: Consider the matrix equation alXl + a2 ,(2 + a3 X3 = 0
i.e al(l, 0,3,1) + a2 (0, 1, 6, 1) + a3 (0, 2, 1, 0) =0
(al + Oa2 + 0 a 3, 0 a 1 + a 2 + 2 a 3, 3 a 1  6 a 1  6 a 2 + a 3, a 1  a 2 + Oa 3) =0
=0
a2 + 2a3 =0
3al  6a2+ a3 =0
and al  a2 =0
ie
158
i.e. al =0, 0,2 + 2a2 =0 => a2 =0
13 a3 =0 => a3 =0
i.e. al =0 = a2 = a3
Matrices
R3 R3  3RI
R4 R4  RI
R3 R3 + 6R2
R4 R4 + R2
Thus, the given vectors are linearly independent. Answer.
Example 13 : Examine the following vectors for linear dependence and find the
relation if it exists.
Xl = (1, 2, 4), X2 = (2, 1, 3), X3 = (0,1,2) and M = (3,7,2)
(U.P.T.U.2002)
Solution: The linear combination of the given vectors can be written in matrix
equations as
al Xl + a2 X2 + a3 X3 + a4 X4 = 0
=> al (1,2,4) + a2 (2, 1, 3) + a3 (0, 1, 2) + a4 (3, 7, 2) =0
=> (al+2a2+0a33a4,2ala2 +a3+7a4,4al+3a2+2a3+2(4)=0
=> al + 2 a2 + 0 a3  3 a4 =0
2ala2+a3+7a4=0
4 a 1 + 3 a 2 + 2 a 3 + 2a 4 = 0
This is a homogenous system i.e.
l!
Applying elementary row transformations we have
l =:
159
A Textbook of Enzj.neering Mathematics Volume  I
[ ~
2
5
o
Hence the given vectors are linearly independent
i.e.a1+2a23a4=0
 5 a 2 + a 3 + 13 a 4 =0
a3+a4=0
putting a4 = k in a3 + a4 =0 we get a3 =k
 5 a 2  k + 13 k =0
. 12 12
I.e. a 2 =  k and a 1 + 2 x  k  3k =0
5 5
9
=>a
1
=k
5
Hence the given vectors are linearly dependent substituting the values of a in
a 1 Xl + a 2 X2 + a 2X2 + a 4 X4 =0 we get
9k 12k
X +x2kX +kX =0
5 1 5 3 4
or 9X1  12 X2 + 5N  5X4 =0
Characteristic Equation and Roots of a Matrix:
Let A = [aij] be an n x n matrix,
(i) Characteristic matrix of A :  The matrix A  AI is called the characteristic
matrix of A, where I is the identity matrix.
(ii) Characteristic polynomial of A: The determinant I A  AI I is called the
Characteristic polynormial of A.
(iii) Characteristic equation of A : The equation I A  AI I =0 is known as the
characteristic equation of A and its roots are called the characteristic roots or
latent roots or eigenvalues or characteristic values or latent values or paper
values of A.
The Cayley  Hamilton Theorem: Every square matrix satisfies its characteristic
equation i.e if for a square matrix A of order n,
I A  AI I = (l)n [An + a1 A
n
1 + a2 An2 + ................... + an]
Then the matrix equation
Xn + a1 Xn1 + a2 Xn2 + a3 X
n

3
+ ...................... an I = 0
is satisfied by X = A
i.e. An + a1 An1 + a2 An2 + .................. + a1 I =0
(U.P.P.C.S. 2002; B.P.Sc 1997)
160
Matrices
Proof: Since the element of A AI are at most of the first degree in A, the elements
of Adj (A  AI) are ordinary polynomials in A of degree n lor less.
Therefore Adj (A AI) can be written as a matrix polynomial in A, given by
Adj (A  AI) = BOA
n

1
+ B1A
n

2
+ ................. +Bn2A + Bn1 where Bo, B1 .............. B
n
1
are matrices of the type n x n whose elements are functions of ail's
Now (A 1.1) adj. (A  AI) = I A AI I I
:. A adj A = I A I In
:. (A  AI) (Bo A
n

1
+ B1An2 + ............ +B
n

2
A + B
n
1)
= (l)n [An + a1A
n

1
+ ............... + an] I
Comparing coefficients of like powers of A on both sides, we get
 I Bo = (l)n I
ABo  181 = (l)n a1 I
ABl  182 = (l)n al I
AB
n
1 = (l)n an I
Premultiplying these successively by An, An1 ................... .1 and adding we get
0= (l)n [An + al An1 + a2 An2 + .............. + anI]
Thus,
An + al An1 + a2 An2 + ................. + an1 A + anI =0 Proved.
Example 14 : Find the characteristic equation of the matrix
A=[; ~ ~ ;]
and verfy that it is satisfied by A and hence obtain A1.
Solution: We have
IAAlI=[ ~ t 2 ~ ~ A ; ~ J
= (2  A) { (2  A)2 I} + 1{1(2  A) +1} + {1 (2  A)}
= (2  A) (3  4 A + A 2) + (A  1) + (A 1)
= A3+6A29A+4
we are now to verify that
A3  6A2 + 9A  41 = 0 (i)
we have
161
(V.P.P.C.S1997; V.P.T.V. 2005)
A Textbook ofEllgineering Mathematics Volume 1
;
o 1 1 1 2
A
2
= A x A [:  :  ]
5 5 6
A
3
=A2
21  21 22
Now we can verify that A3  6A2 + 9A  41
]+! :]+9[; J4[i n
n
Multiplying (i) by A 1, we get
A2  6A + 91  4AI =0
1
:. AI =  (A2  6A + 91)
4
Now A2  6A + 91
=[_: : ]+[612_ 126 :6
5  5 6 6 6  12 0 0 9
1 1 3
:.A
1
1
1
3
1
Answer.
Example 15: Use Cayley  Hamilton theroem to find the inverse of the following
matrix
162
Matrices
A
1 2 1
and hence deduce the value AS  6Al + 6A3  11A2 + 2A + 3
Solution: The characteristic equation of the given matrix is
I A AJ I = 0
[
4; A 1 A _ 12 ] = 0
1 2 1 A
i.e. (4  A) [(1  1..)2 + 4]  3 [2 (1 A) + 2] + 1[4  (1 A)] =0
or 1..
3
 61..
2
+ 61.. 11 =0
By Cayley  Hamilton theorem we have
AL 6A2 + 6A  11 I = 0
Multiplying by Al, we have
Al A3  6Al A2 + 6Al A 11 Al I =0
or A2  6A + 61  11Al =0
11 Al = A2  6A + 61
(U.P.T.U.2002)
=[; :][; : ]6[ : !
=[ ]_[ 1:
9 7  2 6 12 6 0 0 6
= [! 1
3 5 2
Therefore
_\ 1 [ 5
A = 11 :
Now
1
3
5
71
10
2
AS  6A4 + 6A3 l1A2 + 2A + 3 = A2 (A3  6A2 + 6A 111) + 2A + 31
= A2 (0) + 2A + 31
163
A Textbook of Engineering Mathematics Volume  I
=2l : 1
=[; ; ; 1
; 1 Amwer.
Eigen vectors of a Matrix:
Let A be an n x n square matrix, A be the scalar called eigen values of A and X be
the non zero vectors, then they satisfy the equation
AX=AX
or [A  IA] X =0
For known values of A, one can calculate the eigen vectors.
Eigenvectors of matrices have following properties.
1. The eigen vector X of a matrix A is not unique.
2. If Al, A2 ............ An be distinct eigen values of n x n matrix then the
corresponding eigen vectors Xl, X2 ................. Xn form a linearly independent
set.
3. For two or more eigenvalues, it mayor may not be possible to get linearly
independent eigenvectors corresponding to the equal roots.
4. Two eigenvectors Xl and X2 are orthogonal if X1X2 =0
5. Eigenvetors of a symmetric matrix corresponding to different eigenvalues
are orthogonal.
In this section we will discuss four cases for finding eigenvectors, namely.
1. Eigen vectors of nonsymmetric matrices with non  repeated eigenvalues.
2. Eigenvectors of nonsymmetric matrix with repeated eigenvalues.
3. Eigenvectors of symmetric matrices with non  repeated eigenvalues.
4. Eigenvectors of symmetric matrices with repeated eigenvalues.
Example 16 : Find the eigenvalues and eigenvectors of the matrix
:1
(lAS 1994; U.P.P.CS 2005; U.P.T.U. (CO.) 2002)
Solution: The characteristic equation of the given matrix is
164
I A AI 1=0
31.. 1
o 21..
4
6 =0
o 0 51..
or (3  A ) (2  A) (5  A) =0
:. A = 3, 2, 5
Matrices
Thus the eigenvalues of the given matrix are
Al = 2,1..2 = 3, 1..3 = 5
The eigenvectors of the matrix A corresponding to 1..= 2 is
[A  All] X = 0
:
o 0 52 X3 0
:l[:J[]
Le. Xl + X2 + 4X3 =0
6X3 =0 => XCI =0
Xl + X2 =0
=> Xl =  X2 = kl (say) , kl ::F 0
Thus, the corresponding vector is
X,
The eigenvector corresponding to eigenvalue A2 = 3
[A  1.2 I] X = 0
[
:
o 0 5  3 X3 0
l
Le. X2 + 4X3 = 0
X2 + 6X3 = 0
and 2X3 = 0 => X3 =0
165
A Textbook of Engineering Mathematics Volume  I
i.e. X2 =0 (:. X3 =0)
Now let Xl = k2, we get the corresponding eigenvector as
X, =[
Again when A = 5, the eigenvector is given by
[A Ad] X = 0
[
2 :
o 0 55 X3 0
i.e. 2 Xl + X2 + 4X3 =0
 3X2 + 6X3 =0
or X2 = 2X3 = k3 (say), k3 "* 0
Then
2Xl = X2 + 4X3 = k3 + 2k3
= 3k3
3
Xl = k3
2
Thus, the corresponding vector is
X, =[:J ::,'
2 3
Example 17: Find the characteristic equation of the matrix
[
: j
1 2 1
(U.P.P.C.S 2000; U.P.T.U. 2007)
Also find eigenvalues and eigenvectors of this matrix.
Solution:
The characteristic equation of the matrix is
166
Matrices
I A AI I X =0
[
]=0
1 2 21..
=> 1..
3
 5 1..
2
+ 81..  4 =0
=> (A 1) (1..
2
 41.. + 4) =0
or 1..= 1, 2, 2
The eigenvectors corresponding to Al = 1 is
[A 11] X =0
1 2 21 X3 0
2X2 + 2X3 =0
X2 + X3 = 0 => X2 =  X3 = kl (say), Kl =t. 0
 Xl + 2X2 + X3 =0
Xl = 2X2 + X3 = 2kl  kl = kl
Hence, the required eigenvector is
Now, for the eigenvector corresponding to 1..2 = 2
[A  21] X =0
[
1 2 2
2
=
1 2 2  2 x, 0
i.e.  Xl + 2X2 + 2X3 =0
X3 =0
 Xl + 2X2 =0
Xl = 2X2 = k2 (say), k2 =t. 0
167
A Textbook of Engineering Mathematics Volume  I
Diagonalization of Matrices:
We have referred to the elements au of a square matrix as its main diagonal
elements. All other elements are called off diagonal elements.
Diagonal Matrix:  Definition
A square matrix having all off  diagonal elements equal to zero is called a
diagonal matrix.
We often write a diagonal matrix having main diagonal elements dl,
d2 ................. d
n
as
[
d
1
01
d
2
. d
n
with in the upper right and lower left corners to indicate that all off  diagonal
elements are zero.
Diagonalizable Matrix: An n x n matrix A is diagonalizable if there exists an
n x n matrix P such that pl AP is a diagonal matrix.
When such P exists, we say that P is diagonalizes A.
Example 8 : Diagonalize the following matrix
A =[  ~ ~ l
Solution: The eigenvalues of the given matrix are 1 and 3, and corresponding
eigenvectors [ ~ ] and [ ~ ] respectively
From
P = [ ~ ~ ]
Because the eigenvectors are linearly independent, this matrix is non singular
(as I P I =F 0), we find that
p
1
= [ ~  ~ l
168
Matrices
Now compute
P1AP=
[
1 1][1
010
=[  ~ ~ ]
which has the eigenvalues down the main diagonal, corresponding to the order
in which the eigentvectors were written as column of P.
If we use the other order in writing the eigenvectors as columns and define
Q = [ ~ ~ ]
then we get
Q1AQ = [3 0]
o 1
Example 19 : Diagonalize the matrix
[
~ ~ ~ j
o 0 3
(U.P.T.U.2006)
Solution: The characteristic equation of the given matrix is I A  AI I = 0 i.e.
[
1 ~ A 2 ~ A ~ j = 0
o 0 31..
1..= 3,4,1
Now, eigenvectors corresponding to
=> A = 1 is
[A  All] Xl =0
[
1 +1 6
i.e. 1 2+ 1
o 0
6
3
o
i.e. 2Xl + 6X2 + X3 =0
Xl + 3X2 = 0
4X3 =0 => X3 =0
169
A Textbook of Engineering Mathematics Volume  I
Xl =  3X2
suppose X2 = k, then Xl =  3k, k i= 0
The eigenvector is
X, =[::H 3;]=fr]
Eigenvector corresponding to 1.2 = 3 is
[A  A.21]X2 =[)
[
13 6 1 1[XI] [0]
i.e. 1 2  3 0 x
2
= 0
o 0 3  3 X3 0
[
2 6
=> 1 1
o 0
i.e. 2 Xl + 6X2 + X3 =0
Xl  X2 = 0 XI= X2 = k (say), k i= 0
X3 = 2XI  6X2 = 2k  6k =  4k
The eigenvector is
Eigenvector corresponding to 1.3 = 4 is
[A  1.31] X, =0
o 0 3  4 X3
[
3 6
=> 1 2
o
:.  3XI + 6X2 + X3 =0
Xl  2X2 = 0 => Xl = 2X2
 X3 =0 => X3 =0
Let X2 = k then Xl = 2k, Thus, the eigenvector is
170
x, = [ :: H F] = k l
Thus, modal matrix P is
Now p
1
= ;0 [  
4 12 4
For diagonalization
D = PlAP
Matrices
1[4 8 1][1 6 1][3 1 2]
=  0 0 5 1 2 0 1 1 1
20 4  12  4 0 0 3 0  4 0
24 16+0
=_2 0+0+0
20
[
48+0
412+0
 24  24 + 0  4 + 0 12 0  4 0
=
16 48
[
12 +8+ 0
=_2 0+0+0
20 4848
20 0 0
16 0 4 0
4+84
0+060
1648+64
0][1 0]
0
D = dia (1, 3, 4)
Complex Matrices:
Answer.
8+8+0 ]
0+0+ 0
3248+0
A matrix is said to be complex if its elements are complex number. For example
A=[2+3i 4.i]
2 1
is a complex matrix.
171
A Textbook of Engineering Mathematics Volume  I
Unitary Matrix:
A square matrix A is said to be unitary if
A
8
A = I
Where A
8
= (A)" i.e transpose of the complex conjugate matrix.
Example 20 : Show that the matrix
~ [ 1. 1 + i] is unitary
,,3 11 1
Solution:
~ ]
~ ] = I
=> A is unitary matrix. Proved
Hermitian matrix:
1i]
1
l+i]
1
l(l+i)+(l+i).l 1
(li)(l+i)+(I)(I)
(U.P.T.U.2002)
A square matrix A is said to be a Hermitian matrix if the transpose of the
conjugate matrix is equal to the matrix itself i.e
N = A => a;j = aji
where A =[aij]nxn; aij EC
For example
[
a b + iC] [1 .
1 2 + 31
bic d
34i
23i
o
4+Si
3+4i 1
4Si
2
172
Matrices
are Hermitian matrices.
if A is a Hermitian matrix, then
ail = aii ~ a  i ~ = a + i ~
~ 2 i ~ =0
~ ~ = O
:. aii = a. + i (0)
ali = a.
= which is purely real
Thus every diagonal element of Hermitian Matrix must be real.
(U.P.P.C.S 2005)
Skew  Hermitian Matrix:
A square matrix A is said to be skew  Hermitian if A e =  A ~ aij = aij
For principal diagonal
j = i
~ all + aii =0
~ real part of aii =0
~ diagonal elements are parely imaginary
Thus the diagonal elements of a skew  Hermitian matrix must be pure imaginary
numbers or zero.
For example
[ 2 ~ i  2 ~ i J [ _ ~ i+ 4i 3 ~ 4i 1
are Skew  Hermitian matrices.
Problem set
Exercise
1. Using elementary transformations find the inverse of matrix A where
2
4
5
Ans . A 1 = r; : : ~ j :
1/4 1/4
[U.P.T.U. (C.O.) 2007]
~ l
2. Using elementary transformations, find the inverse of the matrix A where
173
A Textbook of Engineering Mathematics Volume  I
: ;1
Anso A
1
=[: :
1 0
30 Using elementary transformations, find the inverse of the matrix A where
1 0 1 J
1/4
(3/4)i
1/4
i / 2]
i / 2
1/2
40 Using elementary transformations, find the inverse of matrix A where
f2
60
12
12
12
]
50 Find the rank of the following matrix by reducing it to normal formo
(i)
f i
(UoPoToD. (C.O) 2002)
(ii)
(UoPoToUo (C.Oo) 2007)
174
Matrices
(iii)
[ ~
3 1
11
1 2 4
1 3 2
3 0 7
(iv)
Find the rank of the matrix, A = [ ~
two non  singular matrices P and Q
such that P AQ = I Hence find A1.
Ans. P=[ ~  ~ ~ land Q[ ~  ~
2 3 3 0 0
and AI = QP = [  ~ ;  ~ 1
2 33
(UP.T.U.2006)
2 3
3 5
3 4
(UP.T.U. 2000,2003)
(iv) 2
(U.P.T.U 2002)
~ 1
7. Find the non singular matrices P and Q such that PAQ in the normal form
for the matrices below.
(i)
(ii)
2
2
o
3
1
4
175
A Textbook Mathematics Volume  I
p=[
0
1/3 4/3
1
/
3]
Ans. (i) 1
1/6 5/6 7/6
0 1 0
1 1
0 0 1
1 1
0
j28J
(ii) P= 0 1
3/28 13/28
3 7
21/28]
1 2 10/28
0 1
47 {28
0 0
8. Use elementary transformation to reduce the following matrix A to upper
triangular form and hence find the rank A where
A
=  = = :]
3 1 3 _ 2 (U.P.T.U.2005)
6 3 0 7
Ans.3
9. Check the consistency of the following system of linear nonhomogenous
equations and find the solution, if it exists.
7Xl + 2X2 + 3X3 = 16; 2Xl + 11 X2 + 5X3 = 25, Xl + 3X2 + 4X3 = 13
(U.P.T. U. 2008)
95 100 197
Ans. x =  x =  x =
I 91' 2 91' 3 91
10. For what values of A and /!, the following system of equations
2x + 3y + 5z =9 , 7x + 3y  2z =8, 2x + 3y + AZ =/!
will have (i) unique solution and (ii) no solution
Ans. (i) A :;t 5 (ii) /! :;t 9, A = 5
11. Determine the values of A and /! for which the following system of
equcrtions
3x  2y + Z = /!, 5x  8y + 9z = 3, 2x + Y + AZ = 1 has
(i) Unique solution (ii) No solution and
(iii) Infinite solutions.
Ans. (i) A :;t 3 (ii) A = 3, /!:;t (iii) A = 3, /! = 1/3
176
Matrices
12. Verify the Cayley  Hamilton theorem for the following matrices and also
find its inverse using this theorem
(i) A" : :]
(V.P.T.V.2007)
(ii)
(V.P.T.V. (CO) 2007)
Ans. (iW (ii) A' _!
13. Find the characteristic equation of the symmetric matrix
i]
and hence also find Al by Cayley  Hamilton theorem. Find the value of A6
6A5 + 9A4  2A3  12A2  23A  91
o 1]
3 1
3 3
l
84
and Value = 80
102
102 80]
106 80
138 106
14. Find the characteristic equation of the matrix
11
II 1 2
(V.P.T.V. 2003, 2004)
verify CayleyHamilton theorem and hence evaluate the matrix equation
AS  SA7 + 7 A6  3A5 + A4  SA3 + 8A2  2A + I
(V.P.T.V.2002)
Ans. AL S/.,2 + n  3 = 0
177
A Textbook of Engineering Mathematics Volume  I
: ~ ]
Find the eigenvalues and corresponding eigen vectors of the following
(i)
[! i ~ ]
[
 ~ ~ :]
1 2 0
(ii)
(iii)
H : :]
Am. (i)1, 1, 7; [J U] [ ~ ]
(ii)
5,3,3; [ _ ~ ] , [ ~ ] [ ~ ]
(in) 2,2,3, [ =m =;]
Dependent eigenvectors.
16. Diagonalize the given matrix
(i) [ ~  ; ]
(ii)
[1
5
33]
(ill) [ ~ ~ _ ~ ]
178
Matrices
0 0
J
4 1
(iv)
0 3
0 1
[3+F7i
Ans. (i) D = P1AP = 0
2
3:F7i I
(ii)
(ill) P'AP = l
0
5
0
1 0 0 0
0 4 0 0
(iv)
P1AP=
0 0
5+Fs
0
2
0 0 0
5Fs
2
17. Prove that is not diagonalizable.
OBJECTIVE PROBLEMS
Each of the following questions has four alternative answers, one of them is
correct. Tick mark the correct answer.
1. The rank of the matrix A = l
2
(A)
(C)
o
2
(B)
(D)
2. The equations
xy+2z=4
3x + Y + 4z = 6
x + Y + Z = 1 have
1
3
3 2]
9 6 is
6 4
(U.P.P.C.S.1990)
Ans. (B)
179
(A)
(C)
A Textbook o(Engineering Mathematics Volume  I
Unique solution
No solution
(B)
(D)
Infinite solution
None of these
Ans. (B)
3. ~ ~ 1 a n d I be 3x3 unit matrix. If M = I  A, then rank of
o 1
I  A is
(A)
(C)
o
2
(B)
(D)
1
3
4. If peA) denotes rank of a matrix A, then p(AB) is
(A) peA) (B) pCB)
(C) Is less than or equal to min [peA), pCB)]
(D) > min [p (A), p(B)]
5. If 3x + 2y + Z =0
x + 4y + Z = 0
2x + Y + 4z = 0
be a system of equations then
(A) It is inconsistent
(B) It has only the trival solution x =0, y =0, z = 0
(LAS. 1994)
Ans. (B)
(LAS. 1994)
Ans. (C)
(LAS. 1994)
(C) \. It can be reduced to a single equation and so a solution does not exist.
(D) The determinant of the matrix of coefficient is zero.
6. Consider the Assertion (A) Reason (R) given below:
Assertion (A) the system of linear equations
x  4y + 5z =8
3x + 7y  z =3
x + 15y l1z = 14 is inconsistent.
Ans. (B)
Reason (R) Rank peA) of the coefficient matrix of the system is equal to 2, which is
less than the number of variables of the system.
(LAS. 1993)
The correct answer is 
(A) Both A and R are true and R is the correct explanation of A
(B) Both A and R are true but R is not a correct explanation of A
(C) A is true but R is false.
(D) A is false but R is true.
180
Matrices
7. Consider Assertion (A) and Reason (R) given below:
Assertion (A) : The inverse of [: :] dose not exist
Reason (R) : The matrix is non sigular.
The correct answer is __
(A) Both A and R are true and R is the correct explanation of A.
(B) Both A and R are true but R is not a correct explanation of A
(C) A is true but R is false.
(0) A is false but R is true.
Ans. (B)
(I. AS. 1993)
Ans. (0)
8. If A = Oiag (1\.1,1.2, ........... An), then the roots of the equation det (A  xl) =0
are_
(A) All equal to 1
(B) All equal to zero
(C) A"l:S;i:S;n
(0) Ai, 1:s; i:s; n
9. The matrix ;] then inverse matrix is given by
2 11
(A)
1/4 3/4
(B)
1/2 1/2]
3/4 5/4
1/4 3/4
(C) ;:: 5
1
/4
2
]
1/4 3/4
(0)
1/4 3/4
181
(I. AS. 1993)
Ans. (C)
Ans. (B)
A Textbook of Engineering Mathematics Volume  I
10. The rank of the matrix A = l: : ~ 1 given by
(A) 0 (B) 1
(C) 1 (D) 3
Ans. (B)
11. The value of A for which the system of equation x + 2y + 3z = Ax, 3x + Y +
2z = AY, 2x + 3y + Z = AX, have non  trival solution is given by
(A) 1 (B) 2
(C) 4 (D) None of these
Ans. (D)
12. Let the matrix be A = [ ~ ; ]and B = [ ~ 1 ~ ] , which one of the
following is true.
(A) AI exists
(C) AB = BA
13. The equations
x + y + z =3
x + 2y + 3z = 4
2x + 3y + 4z = 7
have the solution 
(A) x =2, Y = 1, z =1
(B)
(D)
(B) x =1, Y = 2, Z = 1
(C) x = 3, Y = 1, Z =1
(D) x =1, Y =0, Z = 3
B
1
exist
None of these
Ans. (D)
Ans. (C)
14. If A and B are square matrices of same order, then which one of the
following is true.
(A) (AB)' = A' B' (B) (ABfl = Xl B
1
(C) (AI), = (A'fl (D) B'AB = BA'B
Ans. (C)
I
I 0 01
15. The inverse of the matrix 0 2 0 is
o 0 3
182
1 0 0
(A) 0
1
0 (B)
2
0 0
1

3
1
0 0 
2
(C) 0 1 0 (D)
0 0
1

3
16. The rank of the matrix l  ~
1
(A)
(C)
o
2
(B)
(D)
1
3
Matrices
1 0 0
0
1
0 
3
0 0
1
2
N one of these
Ans. (A)
2]
~ is
Ans. (B)
17. Let A be an n x n matrix from the set of real numbers and A3  3A2 + 4A 
61 =0
where 1 is n x n is unit matrix if Al exists, then
(A)
(B)
(C)
(D)
AI = AI
AI = A + 61
AI = 3A  6 1
Al = .! (AL 3A + 41)
6
(LAS. 1994)
Ans. (D)
18. Consider the following statements. Assertion (A) : If a 2 x 2 matrix,
commutes with every 2x2 matrix, than it is a scalar matrix.
Reason (R) : A 2 x 2 scalar matrix commutes with every 2 x 2 matrix
of these statements
(LAS. 1995, 2007)
(A) Both A and R are true and R is the correct explanation of A.
(B) Both A and R are true but R is not a correct explanation of A
(C) A is true but R is false.
(D) A is false but R is true.
183
A Textbook of Enzineerinz Mathematics Volume  I
Ans. (B)
Ans. (D)
20. The system of equation
x+2y+z=9
2x + y + 3z =7
can be expressed as
(LA.S. 1995, 2004)
2
J
(A)
1
2
(B)
1
(C)
G :
(D) None of these
Ans. (D)
21. The points (Xl, Yl), (X2, Y2), (X3, Y3) are collinear if the rank of the matrix.
[
Xl Yl 1J
A = x
2
Y2 1 is
X3 Y3 1
(A) 1 or 2 (B) 2 or 3
(C) 1 or 3 (P) 2
Ans. (A)
184
Matrices
22. To convert the Hermitian matrix into. Skew Hermitian one, the Hermitian
matrix, must be multiplied by
(A) 1 (B) i
(C) i (0) None of these
l  ~
i+1
5 ~ }s a 3 x 3 matrix over the set of
23. 1
Ji
3i
(A) Natural numbers (B)
(C) Real numbers (0)
24. The system of equations
x + 2y + 3z == 4
2x + 3y + 8z == 7
x  Y + 9z == 1 have
(A) Unique solution (B)
(C) Infinite solution (0)
Integers
Complex numbers
No solution
None of these
Ans. (B)
Ans. (0)
Ans. (C)
25. A system of equation is said to be consistent if there
exist ............... solutions for the system 
(A) No (B) One
(C) At least one (0) Infinite
Ans. (C)
26. If the determinant of coefficient of the system of homogeneous linear
equation is zero, then the system have 
(A) Trivial solution
(B) Non  trivial solution
(C) Infinite solution
(0) None of these
Ans. (B)
27. Who among the following is associated with a technique of solving a
system of linear squations ?
(A) Sarrus (B)
(C) Cramer (0)
Matrix A = r ~
2
28. 0
i
(A) Unitary
Cayley
Hermite
:}s
(B) Hermitian
185
Ans. (C)
A Textbook of Engineering Mathematics Volume  I
(C)
29.
(A)
(C)
30.
(A)
(C)
SkewHermitian
[ 1
Given matrix 0
3 +i
Hermitian (B)
Unitary (D)
Mamx [i
1
0
2i
Hermitian (B)
Unitary (D)
(D) None of these
0
3i]
1 o is
0 0
Non  Hermitian
None of these
~ +
SkewHermitian
None of these
31. Characteristic roots of the matrix
A= IS
[
COS 8  sin 8] .
sin8 cos8
(A) i (B) 1
(C) 2 (D) None of these
32. The eigen values of the matrix
A = r  ~ ~ = !] is
1 2 0
(A) 5, 3, 3 (B) 5,3,3
(C) 5,3, 3 (D) None of these
33. The matrix A is defined by
A = r~ ~ ~ ] the eigenvalues of A
2
is
142
(A) 1, 9, 4 (B)
(C) 1, 3, 2 (D)
1,9,4
1,3, 2
186
Ans. (B)
Ans. (A)
Ans. (B)
Ans. (B)
Ans. (A)
Ans. (B)
34.
(A)
(C)
If the matrix isA ~ [  ~
1,27, 8 (B)
2,50, 10 (D)
Matrices
:] then the eigenvalues of A3 + SA + 81 are
o 2
2
3
 1, 3, 2
2,50,10
Ans. (C)
35. The eigenvalue of a matrix A are 1, 2, 3 the eigenvalues of 31  2A + A2
are
(A)
(C)
36.
(A)
(C)
2,11,6
2,3,6
[ 3i
The matrix  ~
Unitary
(B)
(D)
0
0
i
Skew  Hermitian
3, 11, 18
6,3,11
Hs
(B) Hermitian
(D) None of these
187
Ans. (A)
Ans. (D)
"This page is Intentionally Left Blank"
UNIT4
Multiple Integrals
"This page is Intentionally Left Blank"
Chapter 10
Multiple Integrals
Introduction: The process of integration can be extended to functions of more
than one variable. This leads us to two generalizations of the definite integral,
namely multiple integrals and repeated integrals. Multiple integrals are definite
integrals of functions of several.variables. Double and triple integrals arise while
evaluating quantities such as area, volume, mass, moments, centroids and
moment of inertia find many applications in science and engineering problems.
b
Double Integrals: The definite integral J f(x)dx is defined as the limits of the
sum f(Xl) 8Xl + f(x2) 8X2 + f(x3) 8X3 + ............ + f(xn) 8xn when n ~ ex: and each of
the length 8Xl, 8X2, 8X3 ............... 8xn tends to zero. Here 8XI, 8X2,
8X3 ................ 8xn are n sub intervals into which the range b  a has been divided
and Xl, X2 ............ xn are values of x lying respectively in the first, second,
third, ............ nth subinterval.
A double integral is its counterpart of two dimensions. Let a single valued and
bounded function f(x, y) of two independent variables x, yare defined in a closed
region R of the xy  plane.
Divide the region R into subregions by drawing lines parallel to Coordinate axes.
Number of rectangles which lie entirely inside the region R, from 1 to n. Let (Xr,
yr) be any point inside the rth rectangle whose area is 8A
r
.
y
,,
,.
.....
I
'{
...........
1,...""
o '           ~ X
Consider the sum
f(Xl' YI) 8A + f(x2' Y2) 8A2 + ........... +f(xn, yn) 8An
191
A Textbook of Engineering Mathematics Volume  I
(i)
Let the number of these subregions increase indefinitely, such that the largest
linear dimension (i.e. diagonal) of 8Ar approaches zero. The limit of the sum (i), if
it exists, irrespective of the mode of subdivision is called the double integral of
f(x, y) over the region R and is denoted by
ff f(x,y)dA
R
In other words
n
Lf(x"Yr)8A
r
= ff f(x,y)dA
s'Ar>O R
which is also expressed as
ff f(x,y)dxdy or ff f(x,y)dydx
R R
J
1 f.jl;;! dydx
Example 1: Evaluate 2 2
oOl+x+y
(P.T.U.2006)
Solution: Since the limits of yare functions of x, the integration will first be
performed w.r. t y (treating x as a constant). Thus
r I f.[1;;2 dydx f I [f.[1;;2 dy t
Jo 0 1 + Xl + y2 = 0 0 1 + X2 + y2 rx
= f I [ 1 tan I [ Y )]5.x'dX
o .Jl + Xl .Jl + x
2
0
= r J tanI (l)}dX
o 1 1 + x
2
r,,/2 1 1t
= J, tan
1
tandx
o 1 + x
2
4
1tfl dx
= '4 0.Jl + x
2
= + x2)}I =
Example 2: Evaluate fa1 f:' e
X
/
Y
dydx
Solution: fOI e
X
/
Y
dydx = fo' dy fo'" e x/y dx
192
Answer
Multiple lntezrals
Let t
Y
=> dx = Y dt
= f:ydy{e
y2
/
y
_eo}
= f:ydy(e
Y
1)
= f:y(e
Y
l)dy
= {y(e
Y
y)}:  f>.(e
Y
y)dy
y)): ire' <)t
={l(e
l
1)0}{e
1
o}
1
=e1e+ +1
2
1
2
Answer.
S
dS
J
(d'V
2
) ,r:
Example 3: Evaluate . v
a2
 x
2
 y2 dydx
Solution: Let the given integral be denoted by I
:. I = S: Sf
d2

y2
)  x
2
 y2 )dydx
= f:[S/(d
2

y2
) J{(a
2
_y2)x
2
}dx fY
f:[  y')  x'} +  y' )sin' J(a' y') f'" dy
=   y2) _ (a
2
_ y2)}  y2)sin
1
1 ]d
Y
193
(U.P.P.C.S. 2004)
A Textbook of Engineering Mathematics Volume  I
rl (2 2) 1t d
= J
o
2" a y 2" y
1tfd( 2 2)
="4 0 a y dy
Answer.
6
Example 4: Evaluate Sf (X2 + y2 )dxdy over the region in the positive quadrant for
which x + y::; 1
OR
Evaluate JJ (X2 + y2 ) dxdy , where A is the region bounded by x =0, y =0, x + Y =1.
A
Solution: The region of integration is the triangle OAB, for this region x varies
from 0 to A i.e. from x =0 to x =1 and for any intermediary value of x at N. say y,
varies from the x axis to P on the line AB given by x + Y =1 i.e. y varies from y =0
toy =1x
B (0,1)
(1,0)
~   ~ ~         ~   X
a N A
y=o
194
=
=
= 4x
3
)dx
1 [ 3 2 2 4]1
= 2" x  2" x + 2x  x 0
1%+21]
1
= Answer.
6
Multiple Integrals
Example 5: Evaluate JJ xy dxdy over the positive quadrant of the circle x
2
+
(1 y2)
y2 =1
Solution: The region of integration here is the quadrant OABO of the circle as
shown in figure.
Here, the Coordinates of A and Bare (1, 0) and (0,1) respectively as the radius of
the given circle is 1.
y
B
y= J17
OL__ .. X
y=O
Here the given integral
195
A Textbook of Engineering Mathematics Volume  I
xy
= dxdy
o v=o y2
1 [' JR
= foX v(1y2) 0 dx
= f>[ (1 X2)}  0) JdX
=f>(x1)dx
x)dx
= [*%]
1
= Answer.
6
Example 6: Evaluate If xy(x + Y )dxdy over the area between y = X2 and y = X
(B.PS.C. 2005)
Solution: Here x
2
=y represents a parabola whose vertex is the origin and axis is
the axis of y. The equation y =x is a line through origin making an angle of 45
0
with x axis solving y = x
2
and y = x, we find that the parabola y = x
2
and the line y
=x intersect in the point (0, 0) and (1, 1)
y
Q
X
I X,
Required value = fx=oL=x xy(x + y)dxdy
196
c
1
(1 6 1 7 3 4)
= Jo '2X +"3
X
6
X
dx
= [(2.)x7 + 2. X
8
_ ':!'XS]I
14 24 6 0
1 1 1
=+
14 24 6
= ~ Numerically
56
Multiple Integrals
Answer.
Example 7: Evaluate ff x
2
y2 dx dy over the region x
2
+ y2 :s; 1
Solution:
I
1 IM 2
1= 1 dx _Mx2y dy
I
I 2(1 1)R
 x y'
 I 3 R
= ~ f1 x
2
[(1  x
2
t2 {(1 x
2
t2 JdX
2 J I 2 (1 2 )3/2 d
= x x x
3 I
4 i I 2 (1 2 )3/2 d
= x x x
3 0
= C
n
/
2
sin
2
Scos
3
SdS
3 Jo
putting x = sine
= C
n
/
2
sin
2
Scos
4
SdS
3 Jo
197
A Textbook of Engineering Mathematics Volume  I
1t
Answer. =
24
Example 8: Evaluate If (x + y)2 dx dy over the area bounded by the ellipse
x2 2
+L=1
a
2
b
2
2 2
Solution: For the ellipse .; + ;. = 1 , we have
a b
_X2)
a
(U.P.T.U.2004)
:. Region of integration is for x from a to + a and for y from !? a
2
 X2) to
a
_ X2)
a
f
d f!>.J(d2_x') 2
:. the given integral X=d :=.;J(d'X
2
) (x + y) dxdy
y
p
x =a
x=a
o
= (X2 + y2 + 2xy)dxdy
= 4 I:=o x') (X2 + y2 ) dx dy
Q
(The third integral vanishing as 2xy is an odd function of y)
198
Multiple Integrals
= ;; f: a
2
 X2) [3a
2
x2 + b
2
a
2
 b
2
x
2
Jdx
= f"/2acoaS[3a
4
sin
2
S+b
2
a
2
b
2
a
2
sin
2
SJa cosSdS
3a J6=0
putting x = a sinS
4 f"/2 [ 2 2 2 2 2 J 2
= 3"baJo 3a sin S+ b  b sin S cos SdS
= ba[(3a
2
 b
2
) fO"/2 sin
2
S cos
2
SdS + b
2
fo"/2 cos
2
SdSJ
= ab[(3a'  t
3 2[(3) 2 2
= ab (3a
2
_ b
2
) 2 2 + b 1t
[
!In}Jn 2]
3 2.2 4
= _ b
2
)+ b
2
]
Answer.
Example 9: Find the area lying between the parabola y = 4x  x
2
and the line y =
x.
Solution: The two curves intersect at points whose abscissa are given by
4x  x
2
= x
i.e. x = 0 or 3
199
A Textbook of Enzineerinz Mathematics Volume  I
y
y = 4x  x
2
x
The area can be considered as lying between the curve y =x, y = 4x x
2
, X =0 and x
=3. So, integrating along a vertical strip first, we see that the required area
3 4x_X2
[3 f
4X

x2
f [ ]
=J
o
x dxdy= Y dx
o x
= f:(4xx
2
x)dx
= f:(3xx
2
)dx
2 3
[
3 1]3
= 2"x '3
x
0
= 27 _ 9 = 2. Answer.
2 2
Change to Polar Coordinates:
We have x = r cosS, y = r sinS
Therefore
ax ax
J = ar dS
ay ay
ar as
= Icoss  rSinsl
sinS rcosS
=r
=> If f[ x, Y ]dxdy = If f[rcosS,rsinS]J dSdr
R R'
= Sf f[rcosS,rsinS]rdSdr
R'
Note: In polar dx dy is to be replaced by r dS dr.
200
Multiple Integrals
f
2df&:;!( )
Example 10: Evaluate 0 0 X
2
+ y2 dy dx by changing to polar Coordinates.
Solution: Let 1= r: f):x' (X2 + y2 )dy dx, upper limit of y is
X2 + y2  2ax =0
(x  a)2 + y2 = a
2
y
(i)
x
2
+ y2  2ax = 0
r = 2a cos 8
(x _ a)2 + y2 = a
2
rd8dr
C
This equation represent a circle whose centre is (a, 0) and radius a. Region of
integration is upper half circle. Let us convert the equation into polar Co
ordinates by putting
x = r cos8 and y = r sine
r2  2a r cose =0
r = 2a cose (ii)
:. I = f0
2d
+y2)dydx
= f::02 rao
coss
r2 (rdrde)
f
"/2
= 4a
4
0 cos
4
edO
3na
4
Answer. =
4
I( 2 )
Example 11: Transform the integral J
o
0 y2" X + y2 dx dy by changing to
polar Coordinates and hence evaluate it.
Solution: The given limits of integration show that the region of integration lies
between the curves
y=O,
201
A Textbook of Engineerillg Mathematics Volume  I
Thus the region of integration is the part of the circle x
2
+ y2 =a
2
in the first
quadrant. In polar Coordinates, the equation of the circle is
r2 cos
2
8 + r2 sin
2
S = a
2
i.e. r = a.
y
x=O
o
y=o
x=a
x
y
0=2:
2
Hence in polar Coordinates, the region of integration is bounded by the curves
r =0, r =a, S =0, S = nl2
Therefore,
Cd cVx') I sn/2 fd
JoJo y\,(x
2
+y2)dxdy= 0 )(/2 sin
2
S.r.rdSdr
= fon/2 sin
2
s[ r; I dS
as i"/2 . 2
= sm SdS
5 0
_ as rum
 5 2[(2)
as 1 Jii.Jii 1 S
= =1ta Answer
5 2 2.1 20
Example 12: Evaluate f ~ t
Fx
'] (X2 +y2 )dxdy by changing to polars.
x
Solution: The region of integration is given by y = x, y = ~ ( 2 x  X2), X =0, x =1.
Thus, the region of integration lies between line y = x, a part of circle (x 1)2 + y2
=1, x =0 and x =1.
202
Multiple Integrals
y
x
The diameter of the circle is 2 with its end at (0, 0) and (0, 2). Its equation is r = 2
cos8 and 8 varies from it/4 to 2: (y =x to x =0)
2
Now the given integral in polar Coordinates takes form
f
n
/
2
f2cose r2r d8 dr = .!.fn/2 [r4 J
2
(0,e d8
n/4 0 4 n/4 0
= .!.f7t/2 24 cos
4
8d8
4 7t/4
= 4 cos
4
8d8
f
7t/2
7t/4
f
Tr/2 2
= (1+cos28) d8
7t/4
= fTr/2(1 + 2cos28+ cos
2
28)d8
7t/4
= 1 + 2cos28 + (1 + cos48) d8
f
Tr/2[ 1 ]
Tr/4 2
f
7t/2[3 1]
= +2cos28+cos48 d8
Tr/4 2 2
[
3 1 ],,/2
= 8+sin28+cos48
2 8 ~ 4
3(it it) 1
=  +(01)+(00)
2 2 4 8
= 3it 1
8
Answer.
r
2
r ~ 2 xdydx
Example 13: Evaluate J, J, I by changing to polar Coordinates.
o 0 "x2 +y2
203
A Textbook of Enzineennz Mathematics Volume  I
Solution: In the given integraL y varies from 0 and x varies from 0
to 2.
Y = ..;r
2

X
_
X
2
=> y2 = 2x  x
2
=> x
2
+ y2 = 2x
In polar Coordinates, we have
r2 = 2r cose
=> r = 2cos8
:. For the region of integration r varies from 0 to 2 cos8 and 8 varies from 0 to .:: .
2
In the given integral replacing x by r cos8, y by r sin8, dy dx by r dr d8, we have
1= r1</2 r2cosercos8.r drd8
Jo Jo r
y y
x=o
o
r1</2f2cose
= Jo 0 rcos8drd8
= J."" COSs[ r;r
dS
= r1</22c053 8d8 = 2.3. = i
Jo 3 3
Answer
1t
8= 
2
r = 2cos 8
o
Example 14: Evaluate fo r e by changing to polar Coordinates.
S
' In
Hence show that e
x
dx = 
o 2
(U.P.T.U.2002)
Solution: Given that
1= rfo e(x'+y') dxdy
Here we see that the integration is along a vertical strip extending from y =0 to y
= 00 and this strip slides from x =0 to x = 00.
204
Multiple Integrals
Hence, the region of integration is in the first quadrant, as shown in figure (i)
y t
y
x=o
x ~ 8
    r    ~ ~  ~      + x
o y=o
Fig. (i)
The region is covered by the radius strip from r =0 to r =00 and it starts from e =0
to 8 = n/2 as shown in figure (ii). Thus
y
    ~             ... x
o
1
f (x'+v') f"/2
1
 '
e  dx dy = e r rd8dr
o 0 0 0
= _! r"/2 r (2r ~  r ' drd8
2 Jo Jo
=  ~ fo"/2[ e
r
' I d8
= _!f"/2 (0 1)d8
2 0
= !f"/2
1d8
= ~
2 0 4
Now, let
f
 2
1= o e
X
dx
Also 1= r e
y2
dy
Answer
(1)
(2)
8=0
205
A Textbook of Engineering Mathematics Volume I
(by property of definite integrals)
Multiplying (1) and (2) we get
f = f ~ f ~ e (hY')dxdy = 2:
o 0 4
=> I = ~ ( ~ ) as obtained above
=> I = .Jit Proved.
2
CHANGE OF ORDER OF INTEGRATION
Introduction: We have seen that r r f(x,y) dxdy = r r f(x,y) dy dx, provided
a, b, c, d are constants.
Here we see that the limit of x and y remain the same whatever order of
integrations are performed. In case, the limits are not constant the limits of x and
y in both the repeated integrals will not be the same.
rrf(x,y)dxdy
Means integrate f(x,y) first w.r.t y from y = c to y = d treating x as constant and
then integrate the result obtained w.r.t x from x =a to x =b.
Sometime, the evaluation of an integrated integral can be simplified by reversing
the order of integration. In such cases, the limits of integration are changed if they
are variable. A rough sketch of the region of integration helps in fixing the new
limits of integration.
Note: In some books particularly American (' f,' f(x,y)dydx is written instead
of (' r ~ 2 f (x, y) dx dy where the first integration performed w.r.t y and then after
w.r.t.x.
However in this book we shall generally use the notation given in the beginning
of the introduction.
Example 15: Change the order of integration in the integral
racosuf)(a'x') f(x, y)dxdy
Jo xtdno.
Solution: The given integral rdcosuf)(a'x2) f(x, y)dxdy
Jo xtclnct
Here the limits are given by x =0, x = a cosu; y = x tanu, y = ~ ( a
2
 X2) ,
Y = ~ ( a
2
 X2) gives x
2
+ y2 = a
2
i.e. circle with centre at origin.
To find intersection point of
206
x2 + y2 = a2
and
(i)
y = x tan a (ii)
Multiple Integrals
Now from (i) and (ii) => x
2
+ x
2
tan
2
a = a
2
=>x = a cos a
put this in (ii), y = a cos a tan a
= a sin a
:. (a sin a, a cos a) is the intersection point A of (i) and (ii)
y
x=acosa
y = xtana
Through A draw a line AN parallel to xaxis. Evidently the region of integration
isOAMO.
In order to change the order of integration let us take elementary strips parallel to
xaxis. Such type of strips change their character at the point A. Hence the region
of integration is divided into two parts ONAO, NAMN.
In the region ONAO, the strip has its extermities on the lines x =0 and y = x tan a.
:. limits of x in term of yare from x =0 to
y
 (or y cot a) limits of yare form 0
tan a
to a sin a.
In the region NAMN, the strip has its extremities on the line x = 0 and the circle
X2 + y2 = a
2
:. limits of x in term of yare from 0 to )( a
2
 X2) and limits of yare from a sin a
to a.
:. racosafJ(a2x2) f(x, y)dxdy = fasina ryeota f(x, y )dydx + fa. rJ(a'Y') f(x, y)dydx
Jo xtana 0 Jo as,naJo
Answer.
Example 16: Change the order of integration in the following integrals.
207
A Textbook of Engineering Mathematics Volume  I
(i) r2af:"X <1> (x, y)dxdy
Jo x /4a
(ii) f:ff
a2

x2
)f(x,y)dxdy
Solution: (i) We denote the given integral by I. Here the limits are given by
x =0, x =2a; y = 3ax
x y x
2
or +=1, y=orx
2
=4ay
3a 3a 4a
For intersection point of
y =3a x and X2 = 4ay we have x
2
= 4a (3a x)
or (x + 6a) (x 2a) =0
or x = 2a, 6a
Put in y = 3a x we get y = a, 9a
:. (6a, 9a) (2a, a)
y
x=o
C
x=2a
x
2
= 4ay
x
3a  x = y
Range of integration is OABO. In order to change the order of integration, we
take elementary strip parallel to x axis, such type of strips change their character
at A. Hence we draw line CA parallel to x axis. Thus the range is divided into two
parts OACO and CABC.
In the region OACO, any strip parallel to x axis has its extremities on x =0 and x
2
= 4ay. :. For such strips, limits of x in term of yare from 0 to and limits of
yare from 0 to a.
In region CABC, any strip parallel to x axis has its extermities on x =0 and
x = 3ay and y varies from y =a to y = 3a.
1= fo" <1> (x, y)dydx + f:' f:'
Y
<1> (x, y)dydx
Answer.
208
Multiple Integrals
(ii) Limits x =0, x =a; y =0, y = a
2
 X2) or x
2
+ y2 =a
2
y
(0, a) x
2
+y2=a
2
B 1_ .j
x=a
x=o
o y=o
x
Range is OABO.
:. on changing the order of integration, we get
1= f: f/(a
2
_
y2
) f(x,y)dydx Answer.
Example 17: Change the order of integration of
f2af&
J
o
y)dxdy
Solution: Limits of integration are x =0, x =2a, y =  X2), Y =
Y = jii;; y2 = 2ax, parabola
Y = .J2ax  x
2
+ y2  2ax =0
(x  a)2 + (y  0)2 = a
2
, is circle with centre at (a, 0) and the radius a
==> x = _y2)
intersection of x = 2a and y2 = 2ax is C (2a, 2a).
Range of integration is OABCO. Through A draw a line ED parallel to x axis.
Thus the range is divided in three parts namely.
(1) OAEO
(2) ABDA
(3) EDCE
209
A Textbook o(EnsineerillS Mathematics Volume  I
y
C (2a, 2a)
x = 2a
x=O
==
o (2a, 2a)
A
o
(a, 0) B (2a, 0)
Range No (1) for the region OAEO strip parallel to x axis lies its one end on x =
y2/2a and the other end on x = a  J( a
2
 y2) . For this strip y varies from y =0 to
y=a.
Range No (2) for the region ABDA, the limits for x are froma+J(a
2
_y2) to 2a
and that for yare from y =0 to y =a.
Range No (3), for the region EDCE, the limits for x are from y2j2a (from
parabola) to 2a and that for yare from y =a to y = 2a.
Hence the transformed integral is given by
raJaVll fdf2a f2'f2a
J
o
J(2,x_x,)f(x
y
)dxdy= J
o
)"/2, fdydx + J
o
a+J(d'_I'/dydx+ J
o
//2/
dydx
Answer.
Example 18: Change the order of integration in
J: JJ(:::x') Q>(x, y)dxdy
Solution: We denote the given integral by I. Limits of integration are x =0, x =a;
y = J( a
2
 X2) , Y =x + 2a, y = J( a
2
 X2) gives x
2
+y2 = a
2
circle. y = x + 2a is
expressible as
_x_+L=l
(2a) 2a
210
Multiple Integrals
y
The range is ABCOA. Any strip parallel to xaxis change its character at 0 and C
both. Hence we draw two parallel lines ON and CM. The range is divided in
three parts namely (1) OANO (2) ONMCO, (3) CMBC
Range (1) one end of the strip lies on x = a
2
 y2 ) and the other end on x = a. For
this strip y varies from y =0 to y =a. Similar calculations are done for range (2)
and (3)
f
d fd 5
2d
fa f3d f'
:.1= p,)<l>dydx+ <l>dydx+ <l>dydx Answer.
o vld'y) 0 2. y2d
Example 19: Change the order of integration in rf
e

Y
dxdy and hence find its
J
o
x Y
value.
(LA.S.2006)
Solution: Let 1 = 5f e
I
dxdy
o , y
Here the limits are
x =0, x = 00; y =x, y = 00
y
B 171"::;+ y = 00
x = OpmaulllllllllllllllJP"
211
A Textbook of Engineering Mathematics Volume  I
The range of integration is OABO. In order to change the order of integration, let
us take an elementary strip parallel to x axis. One end of this strip is on x =0 and
the other on x =y. For this strip y varies from y = 0 to y = 00
:. dxdy = dydx
oxy OOy
IV
= dy"dx
o Y 0
y
=
o y
= dy.y
o Y
= eYdy
=
v=o
= 10 =1 Answer.
Example 20: Change the order of integration in f(x, y)dxdy
Jo x' /.
Solution: The limits of integration are given by the parabolas i.e. x
2
/a = y; i.e. X2
a
=ay; xx
2
/ a =y i.e. ax  x
2
= ay and the lines x =0; x = .
2
Also the equation of parabola ax x
2
ay may be written as
this parabola has the vertex as the point its
concavity is downwards.
The points of intersection of two parabolas are given as follows ax  x
2
= X2 or x=
a a a
a,  and hence from X2 = ay, we get y = 0 at x =0 and y =  at x =  .
242
Hence the points of intersection of the two parabolas are (0,0)
Draw the two parabolas x
2
= ay and ax = x
2
 ay intersecting at the point 0 (0,0)
and
2'4
a
Now draw the lines x =0 and x = . Clearly the integral extends to the area
2
ONPLO.
212
Multiple Integrals
Now take strips of the type NL parallel to the x axis.
Solving ay = ax  x
2
i.e x
2
 ax + xy =0 for x, we get
(1)
rejecting the positive sign before square root, Since x is not greater than for the
2
region of integration.
y
x
2
= ay
X
x=o
a
x=
2
Q
Again the region ONPLO, the elementary strip NL has the extermities Nand L
on ax  x = ay and x
2
= ay. Thus the limits of x are from 4ay)Jto
. For limits of y, at 0, y = 0 and at P, y = Hence changing the order of
.. f
a/2
f'" I"
mtegration,wehave 0 ,'/a f(x,y)dxdY=Jo Hd_J(d'_4a
y
)]f(x,y)d
y
dx Answer.
Example 21: By changing the order of integration,
rr y dxdy
o
(LA.S. 2003, V.P.T.V. 2002)
Solution: The given limits show that the area of integration lies between x = y2 j a,
x =y, y =0, y =a since x = y2j a, y2 = ax (a parabola)
213
A Textbook Mathematics Volume  I
and y = x is a straight line, these two intersect each other in the point 0 (0,0) and
A (a, a). The area of integration is the shaded portion in the figure.
(0,0)
X' X
o
Y'
We can consider it as lying between y = x, y = &; X =0, x =a.
Therefore by changing the order of integration we have
i
aJv y faJr.;. ydydx
2 dxdy =
o y 0 x

= rd[_ (ax  y2 t2jVdX dx
Jo a  x
x
(
2 )1/2
f
d ax x
= dx
o ax
= rd(_X )1/2 dx
Jo ax
put x =a sin
2
e
:. dx = 2a sine case de
( )
1/2
1t/2 asin
2
e
= 5 2 2a sine case de
o acos e
5
1t/2 1 n
= 2asin
2
ede = 2a..
o 2 2
na
Answer.
=
2
214
Multiple Integrals
Example 22: Changing the order of integration of fo f o ~ e
xy
sin nx dxdy show that
1
~ sin nx dx = .::
o x 2
(U.P.T.U 2003, 2009)
Solution: The region of integration is bounded by x =0, x = 00, Y = 0, Y = 00.
i.e., the first quadrant, as shown in figure.
y
y=oo
   
 :  

x=oo
x=o
  =.=
  
  

 
   



    ~ ~               ~         ~ x
y=O
Thus
fo fo e xy sin nx dx dy = f o ~ dy fo e xv sin nx dx
=1[ :x)' 2 {YSinnxncosnx}] dy
o n +y 0
= r[o + 2 n 2 ]d
Y
Jo n +y
=[ tan
1
~ I
1t
2
(i) =
on changing the order of integration, we get
f: f o ~ e
xv
sinnxdxdy = r sinnxdx r e
xv
dy
215
A Textbook ofEllzineerinz Mathematics Volume  I
=
o x eX) 0
= sinnx dx[O + 1]
Jo x
= sin nx dx (ii)
Jo x
Hence from equations (i) and (ii) we have
sinnx =
J
o
x 2
Example 23: Find the area enclosed between the parabola y = 4x  x
2
and the line
y = x.
(V.P.T.V.2008)
Solution: The given curves intersect at the points whose abscissas are given by y
= 4x  x
2
and y = x, Therefore
x = 4x  x
2
or 3x  x
2
= 0
=> x (3  x) =0
=> x = 0,3
y
P
y = 4x  x
2
y=x
The area under consideration lies between the curves y = x, y = 4x  X2, X = 0 and x
=3.
Hence, integrating along the vertical strip PQ first, we get the required area as
Area = r
3
J
Y
=4xX'dydx
Jo y=x
216
1
3 [ ]4..
2
= Y dx
o
= f:(4xx
2
x)dx
=
=[%x'
=[%(3)' (Yj
= 27 _ 9 = 2. Answer.
2 2
Multiple Integrals
Example 24: By double integration, find the whole area of the curve
a
2
X2 = y3 (2a  y)
(V.P.T.V.2001)
Solution: The region of integration is shown in figure. Here, we have
Area = 2 x area of the region OAB
= 2f2. ff(Y) dxdy
y=o .=0
y3/2 _ y)
where f (y ) = '....,;.'
a
y
A (0, 2a)
o
Consider the horizontal strip PQ with a small area, we get
'/2 J2ay
Area = = 2f:
a
f: . dxdy
1/, J(2.y)
= 2r2a[x( . dy
Jo 0
= 3. r
2
y3/2  y dy
a Jo
217
A Textbook of Engineering Mathematics Volume  I
putting y = 2a sin28
i.e. dy = 4a sinS cosS dS we get
Area = foll/2(2asin2 St2 4a sinS cosS dS
= 32a
2
fo1t/2 sin
4
Scos
2
SdS
214
x!Jn
= 16a2 2 2 2
6
= na
2
Answer.
EXERCISE
1. Evaluate the integral by changing the order of integration
fo foXxex2/Ydydx
1
Ans. 
2
2. Evaluate
dx f: ey/xdy
1
Ans. 2" (e 1).
(V.P.T.V.2006)
3. Evaluate If xydxdy where R is the quadrant of the circle x
2
+ y2 = a
2
where x
R
o and y 0
a
4
Ans. 
8
4. Evaluate the following integral by changing the order of integration
f1fe dydx
Jo e' log y
Ans. e
1
5. Evaluate by changing the order of integration
flf2 ex'dxdY
Jo 2y
e
4
1
Ans. 
4
218
Multiple Integrals
f
ef,ogYfe'
6. Evaluate 1 1 1 logzdz dx dy
1
Ans.  (e
2
 8e + 13)
4
7. Change the order of integration in
and find its value.
Ans. S: s:' Ix (x + Y )dxdy = f: fav (x + Y )dydx + r f:' II' (x + y)dydx and its value is
8. Change the order of integration of the integral
J: J:/(b+X) f (x, y) dx dy
S
b/a+bsa fl Sb(lY)/Y
Ans. f(x,y)dydx+ f(x,y)dydx
o 0 b/a+b 0
9. Transform f
nl2
f
nl2
~ s ~ n < ! > d<!>de by the substitution x = sin<!> cose, y = sin<!> sine
Jo Jo sme
and show that its value is n.
(V.P.T.V.2001)
10. Let D be the region in the first quadrant bounded by x = 0, y =0 and x + y =1,
change the variables x, y to u,v where x + y = u, y = uv and evaluate
If xy(l x  yt
2
dxdy
o
Ans. 16
945
(V.P.T.V.2002)
11. Determine the area of the region bounded by the curves xy = 2, 4y = X2, Y = 4.
28
Ans.  4 log 2
3
12. Find the volume of the cylindrical column standing on the area common to
the parabolas x = y2, Y = x
2
as base and cut off by the surface z = 12 + Y  X2
219
Ans. 569
140
A Textbook of Engineering Mathematics Volume  I
OBJECfIVE PROBLEMS
(U.P.T.U 2001)
Four alternative answers are given for each question, only one of them is correct.
Tick mark the correct answer.
1. f02 s: (xy + eX fiy dx is equal to
(i) 5 + e
2
(ii) 2(9 + e
2
)
(iii) 2(7 + e
2
) (iv) None of these
Ans. (ii)
2. fo'IY (X2 + y2 ) dy dx is equal to
(
.) 7 (") 44
1  11 
65 105
(
"') 64
1ll 
105
(iv) None of these
Ans. (ii)
3. If R s the region bounded by x =0, Y =0, x + Y =1, then Sf (X2 + y2 )dxdy is equal
R
to
(i) .!.
(") 1
11 
3 5
("') 1
(iv) ..!.. 1ll 
6 12
Ans. (iii)
4. The area bounded by the parabola y2 = 4 ax, xaxis and the ordinates x =1, x =2
is given by
(i)
(ii) 1)
(iii) + 1) .
3 ,.
(iv) None of these
Ans. (ii)
5. The area above the x  axis bounded by the curves X2 + y2 = 2ax and y2 = ax is
given by
220
Multiple Integrals
(iv) None of these
Anso (iii)
60 The area bounded by the curev xy = 4, Y axis and the lines y =1 to Y =4 is given
by
(i) 2 log 2
(iii) 8 log 2
(ii) 4 log 2
(iv) None of these
Anso (iii)
70 The volume of the area intercepted between the plane x + y + Z = 1 and the Co
ordinate planes is
(
0) 1 (00) 1
1  11 
2 3
(
000) 1
111 
6
(iv) None of these
80 rlf"f'+\' xdzdydx is equal to
Jo 1 2\'
(i) ! (ii) 2
6 10
(
000) 1
111 
30
(iv) 2
15
Anso (iii)
Anso (iii)
90 The volume of the tetrahedron bounded by the Coordinate planes and the
plane x + y + Z = 4 is equal to
(
0) 32 (00) 16
1  11 
3 3
100 The volume of
x
2
y2 Z2 0
++=11S
a
2
b
2
c
2
(i) ~ n abe
3
(
000) 6 be
111 na
5
(
000) 4
111 
3
(ii) in abc
3
(iv) in abc
5
(
0 ) 128
IV 
3
Anso (i)
Anso (ii)
221
A Textbook of Ensineerins Mathematics Volume  I
(i) (e 1)3
(iii) (e  2)2
(ii) ~ (e 1)
2
(iv) None of these
Ans. (i)
x y z
12. The volume of the tetrahedron bounded by the plane  +  +  = 1, a, b, c > 0
abc
and the Coordinate planes is equal to
(i) .!.n abc (ii) ~ n abc
3 3
(
... ) 1 be
111 na
6
13. The value of
r1fl, r1'xdzdxdy is
Jo y. J
o
(i) ~
35
(i) ~
2
(
.. ) 3
11 
35
(
.. ) n
11 
4
(iv) None of these
(
... ) 8
111 
35
(
... ) n
111 
3
(iv) None of these
(iv) ~
35
15. The value of the integral f ~ f ~ (X2 + y2 )dxdy is
(i) 1 (ii) 0
(
... ) 1 (. ) 2
111  IV 
3 3
16. The surface area of a sphere of radius r is
(i) 4nr
(iii) 6nr
(ii) 4nr2
(iv) 8nr
2
222
Ans. (iii)
(V.P.P.e.S.1994)
Ans. (i)
(V.P.P.e.S.1995)
Ans. (ii)
(V.P.P.e.S.1995)
Ans. (iv)
(R.A.S 1995)
Ans. (ii)
Chapter 11
Beta and Gamma Functions
Introduction: Beta and Gamma function are improper integrals which are
commonly encountered in many science and engineering applications. These
function are used in evaluating definite integrals. In this chapter we will study
the beta and gamma function and apply them to some common problems.
Definition: The first and second Eulerian integrals which are also called Beta
and Gamma functions respectively are defined as follows
B(m, n) = J ~ x
m

1
(1 X)"1 dx
and I(n) = r e
x
x
n

1
dx
B(m, n) is read as Beta m, n and In is read as Gamma n.
Properties of Beta and Gamma Functions:
Property I: B(m, n) = B (n, m) i.e. Beta functions is symmetrical with respect to m
andn ..
Proof: We know B(m, n)
i
1 m'(l )n1
d = x x x
o
= J ~ (1 x) ml1 (1 X)]"l dx
by the property J: f(x)dx = J: f(a  x)dx
1
1 n1 (1 )ml d
= x x x
o
= B(n, m) Hence Proved
Property II: In = (n l)/(n 1), for all values of n.
Proof: We know that
In = J: exxn1dx
=[xn1(ex)I  Jo(n1)x
n

2
(e
x
)dx
integrating by part taking x
n

1
as first function
=O+(n1)r x
n

2
e
x
dx
= (n l)l(n 1) Hence Proved
Replacing n by (n + I), we get l(n + 1) = nln
223
(u.p.p.eS.1992)
A Textbook ofEllgilleering Mathematics Volume  I
Transformations of Gamma functions:
(i) J
1
dy
Proof: We have
= fo e
X
x"ldx
write e
X
=y Then e
x
dx = dy
 x log e = log y
= log Y
Now the equation (i) reduced to
f(logyt
1
(dy)
IW = f: (lOg r dy Hence Proved
(ii) Prove that
In = k" f: e
kY
y"ldy
Proof: Since we know that by definition of Gamma function
= r e
X
x"ldx (i)
Suppose x =ky then dx = k dy
Now the equation (i) takes the form
r e
ky
(ky)"l kdy
= k" r ekYy"ldy
or I(n) = k" fo e
kY
y"ldy Hence Proved.
Transformation of Beta functions:
(i) Prove that
ml
B (m n) = r y dy
, Jo (1 + y r+"
Proof: Since we know that
B(m,n) = fo\m1 (1 X)"l dx
write x = _1_ then dx = 1 2 dy
l+y (l+y)
with these values (i) becomes
(LA.S. 1998, B.P.S.c. 2007, LA.S. 2005)
(i)
224
Beta and Gamma (unctions Functions
nI
S
 y
or B(m,n) = m+n dy
o (1 + y)
(ii)
Interchanging m and n, observing that B (m, n) = B (n, m) we have
ml
f
 y
B(m,n) = m+n dy
o (1 + y)
(iii)
Adding (ii) and (iii) we get
 ml + nI
2B(m,n) = f y ~ + n dy
o (l+y)
_ x
m

1
+ x
n

1
or 2B(m,n) = I m+n dx
o (l+x)
Relation between Beta and Gamma Functions:
To Prove that
B(m,n)= ~ [ ( n )
I(m+n)
(LAS. 2005, V.P.T.V. 2001)
(LAS. 1990, LAS. 1991, B.PS.C. 1993,97,2005, V.P.T.V. 2001)
Proof: We know that
~ = So e'x
n

1
dx (i)
put x =ay then dx = ady
... ~ = r e
dY
(ay)"I ady
or Un) = re
dy
y
n
1
dy
an Jo 
or I(n) = an S: ea'xn1dx
Pu tting a =y, we get
~ = yn fo e
yx
x
n

1
dx
(ii)
Multiplying both sides by eY yml dy and integrating,
I(n)foe"ym1dy = fo evymIyndyfo e
vx
x
n

1
dx
225
A Textbook of Engineering Mathematics Volume  I
or In 1m = r(r(eY(I+X)ym+nldy)xnldx
orin 1m = f: IX"'dX using (ii)
1m In  x
n

1
or =1 dx
Im+n 0 (1+xr+n
= B(m, n)
1m In
Thus B(m,n) = r Hence Proved.
im+n
Some Important Deductions
(i) To prove that rnl(1 n) = _. _1t_
SInn1t
Proof: We know that
B(m,n)=
I(m +n)
Putting m+n =1 or m = 1 n, we get
B( 1 )
I(i)  n, n
nI
1
 y
we have B(m,n) = ( ) +n dy
o 1+y m
(i)
nl
:.B(n,1n)= 1Ldy=.1t,n<1
o 1+y SInn1t
:. From (i) we have
n) = _. 1t_
SInn1t
Hence Proved.
(ii) To prove that 1(1 + n)I(1 n) =
SInn1t
Proof: Since we know
1t_
SInn1t
Multiplying both sides by n, we get
SInn1t
226
(I.A.S. 1990)
Beta and Gamma (unctions Functions
or
smmt
(iii) To prove that
r" sin
m
o cos" OdO =
o 21( m +;+2)
Proof: we know that
B(p,q) = fOI x
p

I
(1 X)qI dx
f(P)rfq)
= i(p+q)
Hence Proved
putting x = sin
2
S => dx = 2 sinS cos8 d8, we get
x
p

I
(1 X)ql dx = fo"/2(sin2 8r
1
(1 sin
2
st
1
.2sin8cos8d8
= 2J:
/2
sin
2p

1
8COS
2q

1
SdS
1
,,/2 1
:. sin
2p

1
8COS
2q

1
SdS = B{p.q)
o 2
or 1"/2 sin
2p

1
8COS
2q

1
SdS = IPrq
o 2i{p +q)
putting 2p1 = m and 2 q 1 = n
m+1 n+1
or p= and q =
2 2
we get
J"" sin
m
Ocos" OdO = Hence Proved.
o 21(m+;+2)
Legendre's Duplication Formula
To prove that
= 2f:l.r(2m)
Proof: Since we know that
(I.A.S. 1993,1997, U.P.PC.S. 1996, U.P.T.U. 2000)
227
A Textbook of Engineering Mathematics Volume  I
1"/2 sin
2m

1
8COS
2
n
1
8d8 =
o 21(m +n)
(i)
putting 2nl =0 or n =.!., we get
2
1"/2 sin
2m

1
8d8 = Il2")
o 2H
2H
(ii)
putting n =m in (i), we get
(,,/2
JI sin
2m

1
8COS
2
m
1
8d8 =
o . 21(2m)
. 1 "/2. 2ml
I.e. 21 (sm28) 2d8=
2 m 0 21(2m)
putting 28 = q, and 2d8 = dq" we get
1 n
_ r sin
2m

1
q,dq, =
2
2m
Jo 21(2m)
2 1"/2
or  sin
2m

1
q,dq, =
2
2m
0 21(2m)
,,/2 2
2m

1
or 1 sin
2m

1
q,dq, =
o 21(2m)
(iii)
(n/2 2 1
Equating two values of Jo sin m 8d8
from (ii) and (iii) we get
_
2\(
228
Beta and Gamma (unctions Functions
Hence
Example 1: Prove that
1
 x'd Jic
e X=
o 2
Solution: Since we know that
rw= Jo e
t
t
n

1
dt
putting t = x
2
in Gamma function, we get
I(n)= 2r x
2n

1
e
X
' dx
Putting n = .!. we have
2
rn=2Jo ex'dx
or Jic = 2r e
x
' dx
1
 ,' Jic
or e dx= Hence Proved
.a 2
Example 2: To prove that = Jic
Proof: We know that
Inhn::
1t

sinn1t
P
. 1
utting n =  , we get
2
{rnr =n
or i( = In Hence Proved.
Example 3: Show that
B(m, n):: B(m+1, n) + B(m, n+1)
229
Hence Proved
(I.A.S. 1990)
(V.P.P.C.S.1995)
A Textbook of Engineering Mathematics Volume  I
Solution: R.H.S. = B(m+1, n) + B(m, n+1)
I(m + 1).rn + 1)
= I(m+l+n) +#:I(m===+=n=+==l)
ml(m).rn 1m
= (m + n) i( m + n) + :( m+n:) IF=( m=+=n=)
Im.rn [m n]
=I(m+n) m+n + m+n
Im,rn
I(m+n)
= B(m, n)
= L.H.S. Hence Proved
Example 4: Show that
J(e+ 1) e+ 1
Jo eX (loge)
Solution:
i
ooxc
Let 1= dx
o eX
put eX = e
t
x log e = t
t
=> x=
loge
dt
=>dx=
loge
so 1= roo (_t_)r
, Jo loge e
t
loge
=_( 1 )e+1i"" ettCdt
loge 0
= _1_e + 1 foo e
t
t
c
+
1

1
dt
(loge) 0
_M
 (loger
1
Henee Proved.
ExampleS:
Prove that
230
Beta and Gamma functions Functions
)nsecnx
Solution: L.H.5. = x )ITl.( + x )rFx
.: I(n + 1) = nln
n_
smnn
(
1 2) n
= "4
x
)nsec(nx)
Hence Proved
= R.H.S.
Example 6: Show that
Jo"/2 tan" ada = %sec( n;)
Solution: We have
("/
2
r"/2 ( )"
Jo tan" EXia = Jo sin" a cosa da
231
A Textbook of Engineering Mathematics Volume  I
_

 2fi
.: rn/(1 n) = _. _7t_
smn7t
1 7t
= 2: sin l(n + 1)7t
2
= cos ce { (n + 1) 7t }
= cos ce ( + )
= n27t) Hence Proved.
Example 7: Show that
B(n,n) =
2
2n

1
H
Proof: We know that
B(m,n) = x
m

1
(1 X)"l dx
put x = sin
2
e => dx = 2 sine cose de
. ( ) J"12 (. )2m2 ( )2n2
:. B m,n = 0 sme cose .2smH cosede
= 2Jo"/2 sin
2m

1
ecos
2
n
1
ede
/(m+n)
Hence B(m, n) =
/(m + n)
From (i) we have
(i)
232
(V.P.P.C.S.1990)
Beta and Gamma functions Functions
B(n,n) = 2 fo"/2 sin
2n

1
8COS
2n

1
8d8
2 r"/2 2n1
= 2
2n

1
Jo (2sin8cos8) d8
2 i"/2. 2n
1
28d8
= sm
2
2n

1
0
2 1 i" . 2nl"'d'"
=. SIn 'I' 'I'
2
2n

1
2 0
2 1 f"/2. 2nl
=21..2 sm C\>dC\>
2 n 2 0
Example 8: Find the value of
Solution: we know that
jP+ll
q
+
1
1"/2 sin
P
8cos
Q
8d8 = 2 2
o 2l
p
+i +2
putting p = q =0 we have
f
"/2 d8=
o 2f1
=> [e 1;" =
=>
Hence Proved.
233
A Textbook of Engineering Mathematics Volume I
Answer.
Example 9: To show that
r pn)cosnS
(i) Jo edxx
n

1
cosbxdx = I ~ n ) n
(a
2
+b
2
)2
(V.P.T.V. c.C.O. 2004)
(U.P.T.V.2003)
Solution: we know that
reaxxn1dx = I(n) , where a, n are positive put ax =z so that dx = dz
Jo an a
i
 ax n1d S _z(z)nl dz
e X x= e  
o 0 a a
= 2.J ezzn'dz = f(r0
an 0 an
Replacing a by a + ib, we have
Se(d+ib1X xn'dx = ~
o (a + ib)"
Now e (a + iblx = e ax. e ibx
= e ax (cos bx  i sin bx)
Putting a = r cosS and b = r sinS so that
b
r2 = a
2
+ b
2
and S = tan
1

a
(a + ib/ = (r cosS + ir sinS/
= rn (cosS + i sinS)"
= rn (cos nS + i sin nS)
(De Moivre's theorem)
:. From (i) we get
r e
ax
{cosbxisinbx)x"1dx= ~
Jo rn (cosnS+ isinnS)
234
Beta and Gamma functions Functions
= I(n) (cosn8 + isinn8t
rn
= I(n) (cosn8  isinn8)
rn
Now equating real and imaginary parts on the two sides, we get
re"x
n

1
cosbxdx = I(n) cosn8
Jo rn
and r e
dx
x
n

1
sin bxdx = I(n) sinn8
Jo r"
b
where r2 = a
2
+ b
2
and 8 = tan
1

a
Example 10: Prove that
Sf
'I m
1
d d Ii 1m h'+m
D X Y X Y=IZ+m+1
where 0 is the domain x 0, Y 0 and x + Y h
Solution: Putting x = Xh and y = Yh, we get
dx dy = h
2
dX dY
Therefore,
fJ x'l
y
m
1
dx dy= fJ (Xh)'' (Yh)"'l h
2
dX dY
D D'
where 0' is the domain X 0, Y 0, X + Y 1
fJ x'l
y
m
1
dxdy = h'+m f: folXXllymldXdY
D
= h' J:x't.: r dX
h'+m I
=r X
I

1
(lX)"' dX
m Jo
h'+m
=B(l.m+1)
m
h'+m [(Z)1(m + 1)
_ h'+m[(z)mlm
 ml(l+m+1)
235
(V.P.T.V. 2005)
h'+mRl)rnlrn
 l(l+rn+l)
A Textbook of Engineering Mathematics Volume  I
Hence Proved.
1
11/2 de 111/2 ~
Example 11: Prove that .J' x v sin e de = 7t
o sme 0
Solution:
(11/2 de = (1I/2(sinet/2 coso ede
J
o
.Jsine J
o
236
Beta and Gamma functions Functions
Dirichlet's Theorem for Three Variables:
If 1, m, n are all positive, then the triple integral
III
11 m1 n1d d d [[!min
v x y Z x y z=I(l+m+n+1)
where V is the region x;::: 0, y ;::: 0 Z ;::: 0 and x + y + Z 1
Poof: Putting y + Z 1  x = h. Then z h  y
Iff x'l
y
m1
z
n
1
dxdydz = I: x'ldx Io
1

x
ym
1
dy Io
1

X

Y
zn
1
dz
V
= I: x'l dx[Io
h
Jo
h

Y
ym1zn1dydzJ
(put X = h)
= r X'l dX[ !min h
m
+
n
1
o l(m+n+1)
!min r
1
H (1 )m+n d
= JI X x x
l(m+n+1) 0
rxTIln
= Im+n+1
!mrn [[I(m +n + 1)
= \(m+n+ 1) \(1 +m +n+ 1)
(V.P.T.V.2oo5)
IfJ
11 ml n1dxd d [[!min
x y z Y
I(I+m+n+l)
Hence Proved.
N t 1 HI '1 ml n
1
d d d Wmln h/+m+n
oe : v x y z x y z=I(I+m+n+l)
Where V is the domain, x ;::: 0, y ;::: 0 z ;::: 0 and x + y + Z 5 h
Note 2: Dirichlet's theorem for n variable, the theorem states that
III J
',1 '21 I.I d d  ........ h,,+1
2
+ ... +I.
.......... Xl x
2
........... Xn dX
1
X
2
.. .. .... xn 1(1 + 11 + 12 + .... + In)
Example: State the Dirichlet's theorem for three variables. Hence evaluate the
integral HI x/lymlzn1dxdydz where x, y, z are all positive with conditions
(V.P.T.V.2oo5)
'137
A Textbook of Engineering Mathematics Volume  I
Solution: The required integral HI x/lymlznldxdydzwhere the integral is
extended to all positive values of the variables x, y and z subject to the condition
( ~ J + ( ~ J + ( ~ J $1 Let us put
by Dirichlet's integral
pqr
Answer.
2 2 2
Example: Find the mass of an octant of the ellipsoid \ + y 2 + Z 2 = 1 , the density
abc
at any point being p = kxyz
Solution: We know that
Mass = HI pdv
= HI (kxyz)dxdydz
= k HI (xdx) (ydy) (zdz)
(V.P.T.V. 2002, 2006)
(i)
238
Beta and Gamma functions Functions
x2 y2 Z2
Putting =u,= v,=wand u + v + w =1
a
2
b
2
c
2
so that 2xdx = du, 2ydy = dv, 2z?z = dw
a
2
b
2
c
Mass = k Iff ( )( )( )
= Answer.
48
Example: The plane + r + = 1 meets the axes in A, Band C. Apply Dirichlet's
abc
integral to find the volume of the tetrahedran OABC. Also find its mass if the
density at any point is kxyz.
(U.P.T.U.2004)
Solution: The volume of the tetrahedron OABC is given by V = Iff dxdydz for all
D
is positive values of x, y and z subjected to the condition + r + = 1
abc
putting = u =>dx = adu, r = v => dy = bdv and = w => dz = cdw, we get
abc
V = Iff abcdudvdw
D'
where u 0, v 0, w subjected to the condition u + v + W 1
V = abcIff UllV11W11dudvdw
D'
Applying Dirichlet's integral, we get
V = abc 11Hri = abc
1(1+1+1+1) 14
abc
=
6
239
A Textbook of Engineering Mathematics Volume  I
Now mass = HIkxyzdxdydz
D
= k HI au bvcw(abc)dudvdw
D'
= ka
2
b
2
c
2
HI uvwdudvdw
D'
ka
2
b
2
c
2
[2[2[2 ka
2
b
2
c
2
 ?ol (2=+=2=+=2=+ "7
1
) = ;::f7=
= Answer.
720
Example: Evaluate I = HI
V
where V is the region in the first octant bounded by sphere x
2
+ y2 + Z2 =1 and the
Coordinate planes,
[V.P.T.V. (CO.) 2003]
Solution:
Let x
2
= u => x = Fu therefore dx = 1, du, y2 = V Y = .,Jv, therefore
2"u
1 2 'f 1
dy= cdv, z ,dw
2"v 2"w
Then u + v + w =1 Also, u 0, v 0, w 0
1= ffJ(Fu)a1 (rw)V1
v 2Fu2.,Jv2rw
1 Iff L1 L1
=8 u
2
v
2
w
2
dudvdw
Answer.
LIOUVILLE'S EXTENSION OF DIRICHLET THEOREM:
If the variables x, y, z are all positive such the ht< (x + y + z) < h2 then
IH f{x + y + z)x/lymlznldxdydz = rt 1m rn f2 f(u)u
/
+
m
+
n

1
du
1(1 +m+n) h,
Proof: Let I = HI x/1ym1zn1dxdydz under the condition x + y + Z ::; u then
240
Beta and Gamma functions Functions
1= (M + f(m) + rfn))u
/
+
m
+
n
1(1+m+n+1)
(i)
(by Dirichlet's theorem)
if x + Y + Z ::; U + ou, then
1= (u + 8ut
m
+
n
[lim rn
1(l+m+n+1)
(ii)
Now if u< x + y + z < (u + 8u) then
Ilf
11 m1 n
1
d d d [lim rn [( l:: )/+m+n I+m+n]
X y Z x y z = I u + uU  u
(l+m+n+1)
= [l 1m rn u
/
+
m
+
n
[1 + (8u )/+m+n 1]
1(l+m+n+1) u
[l 1m rn I+m+n [1 (1 ) 8u ]
= 1(1+m+n+1) u + +m+n ~ + ......... 1
[l 1m rn I+m+n (1 ) 8u
= u +m+n
1(l+m+n+1) u
[l 1m rn I+m+n l::
= U uU
1(1 +m + n+ 1)
Now consider ffJ f(x + y + z)x/lymlzn1dxdydz. Under the condition hI::; (x + y +
z) ::; hz. When x + y + z lies between u and u + 8u, the value of f(x + Y + z) can
only differ from f(u) by a small quantity of the same order as ou. Hence
ffJf(x+y+z)x/1ym1zn1dxdydz= [lim rn Jf(u)u
/
+
m
+
n

1
0u
1(1 + m +n + 1)
Where x + y + z lies between u and u + ou
Therefore
Jff f(x + y + z)x/lym1znldxdydz [lim rn f2 f(u)u
/
+
m
+
n

1
du
1(I+m+n+1) h,
Example: Evaluate
JJJlog(x + y + z)dxdydz, the integral extending over all positive and zero values
of x, y ,z subjected to x + y + z < 1
Solution: 0 ::; x + y + z < 1.
:. ffJlog(x +y +z)dxdydz = ffJ xHyHzlllog(x+ y +z)dxdydz
241
A Textbook of Engineering Mathematics Volume  I
= [(i) rm rm f e+l+
1

1
log tdt
R3) 0
by Liouvill's extension of Dirichlet's theorem
= '.!. r
J
e log tdt
2 J
o
= logt  i .dt
l[(e }l leI 1
2 3 0 0 2 t
= )}  Answer.
Example: Evaluate
Iff ............. .f dX
I
..... .......
n )1 Xl  X
2
......... xn
integral being extended to all positive values of the variables for which the
expression is real.
(V.P.T.V.2001)
Solution: The expression will be real, if
1  X
l
2  X2
2
 .................  xn
2
> 0
or Xt
2
+ X2
2
+ ................. + Xn
2
< 1
Hence the given integral is extended for all positive value of the variables Xl,
X2 .......... .. Xn such that
o < Xl
2
+ X2
2
+ ................. xn
2
< 1
Let us put Xl2 = UI i.e. Xl = ru; so that, dX
1
= dU
1
etc.
2"ul
Then the condition becomes, 0 < UI + U2 + ............ + Un < 1
1 Iff I U
1
/2U
1
/
2
..........
:. required integral = n ......... du
1
du
2
....... du
n
2 n 1u
l
u
2
.......... u
n
1 Iff I u
2
u
2
..........
= n ......... 2 du
l
du
2
....... du
n
2 n 1u
1
u
2
.......... u
n
= 1 rmm .. rn f_l . ......
2
n
.......... 0
By Liouville's Extension
242
Beta and Gamma {unctions Functions
=2Jmr J1 1
2" rm ofh
1 (fit)" J"/2 1 (
= I .J sin
2
8)2 .2sin8cosd8
2" (n/2) 0 1sin
2
8
putting u = sin
2
8
= 1 (fit)" f"/2
sin
"18d8
2"1 I(n / 2) 0
1 (nt2 rmm
=?I(nj 2) zI("; 1)
Answer.
EXERCISE
1. Evaluate (i) 1(%) (ii) 1(%)
3. Prove that r e
ay
yn1dy = I(n)
Jo an
J
"/2 8
4. Show that sin
3
XCOS
5
/
2
xdx =
o 77
f
1 x2dx
5. Evaluate
0.,.,1_ x
3
243
Ans. (i) 2fit (ii)
3
1
Ans.
280
A Textbook of Engineering Mathematics Volume  I
2
Ans. 
3
6. Express f: xm (1 xn t dx in terms of the beta function, and hence evaluate
f:x5(1x3t dx
8. Prove that I(n + .!.) = J7t1r2n + 1)
2 2
2n
(n + 1)
9. Evaluate
(1 x
6
)
10. Evaluate r x
8
24 dx
o (l+x)
12. Evaluate cosx
2
dx
J
 1t
 2
13. Evaluate
(i) fo xl/4eJXdx
(ii) fo x
2
e _a
2
x
2
dx
244
Ans. (i) .:!.13(m+l,p+l)(ii)
n n 396
1
Ans. 
SODS
1
Ans. '2 log 2n
Ans.O
63
Ans.l
.fit
Ans. 3
4a
Beta and Gamma functions Functions
1 ( 1)"1
14. Evaluate fo x
m

1
dx, m > 0, n >
15. Find the volume of the solid surrounded by the surface
(
X )2/3 (y )2/3 (z )2/3 _
 +  +  1
abc
Ans. In
m"
(V.P.T.V.2008)
Ans.4nabc/35
nabc
Ans. 
6
17. show that Hf dxdydz 3 = .!.log 2  the integral being taken thought the
(x+y+z+l) 2 16
volume bounded by the planes x = 0, y =0, z = 0, x + y + z +1
OBJECTIVE PROBLEMS
Four alternative answers are given for each question, only one of them is correct.
Tick mark the correct answer.
1. fo e'x
3
(i) 2
(ii) 6
is equal to
(ii) 4
(iv) None of these
2. When n is a positive integer then f: ( log "1 dx is equal to
(i) In 1 (ii) In
(ii) I(n + 1) (iv) .!.I(n + 1)
2
3.When n then the value of I(%n) I(%+n}s
(i) n
2
) 1t sec n1t (ii) n
2
) 1t cosec n1t
(ii) n
2
) 1t sin n1t
(iv) None of these
245
Ans. (iii)
Ans. (ii)
A Textbook of Enzineerinz Mathematics Volume  I
4, The value of [(.1)[(.2)[(3) ........ 1(.9) is
(i) 2n
Fa
(
2 )10/2
(ii) Ko
10
(2nt2
(ii) r:;;;
,,10
(iv) None of these
r
l
dx
5, JI r; is equal to
0,,1 x
(i) B (1, 1) (ii)
(iii)B(O,%)
6, B(2, 3) is equal to
(
') 1 (") 1
1 11 
10 12
(
") 1
11 
24
(iv) None of these
r
lt
/
2
7, Jo sin
9
xdx is equal to
(
') 64 (") 64
I  11 
65 35
(
.. ) 64
11 
315
(
' ) 128
IV 
315
8, folt sin
3
x (1 cos x)2 dx is equal to
(
') 2 (") 2
I  11 
3 5
(
") 4 (' ) 8
11  IV 
5 5
9, The value of
r
a
Jo
246
Ans, (i)
Ans, (iii)
Ans, (iv)
Ans, (ii)
Ans, (iv)
Ans, (iv)
Beta and Gamma functions Functions
(i)
2
(
.. ) a
11 
4
(
.. ) na
11 
2
(
. ) na
IV 
4
f"/2 3
10. Jo cos
s
asin ada is equal to
(i) (ii)
32 24
(
.. ) 1 (. ) 1
11  IV 
16 24
2 2
11. Jo e
a
x dxis equal to
(i) fit (ii) fit
2a
(
.. ) fit
11 
4a
(iv)
2a
12. Match the list I with list II
List I
(a) r e
x
x
4
dx
f"/2
(b) J
o
sin
2
xcos
3
xdx
(c)
(d)
The correct match is
a b c
(i) (1) (2) (3)
(ii) (1) (3) (4)
(iii) (2) (1) (3)
(iv) (2) (1) (4)
List II
(1)
15
(2) 24
(3) 5n/256
(4) 12n
d
(4)
(3)
(4)
(3)
247
(M.P.P.CS.1995)
Ans. (iii)
(M.P.PCS.1995)
Ans. (iv)
Ans. (ii)
(V.P.P.CS.1994)
Ans. (iii)
A Textbook of Engineering Mathematics Volume  I
13. Consider the Assertion (A) and Reason given below.
Assertion (A)
J ~ sinxdx = 1 cost
Reason (R) sin x is continuous in any closed interval [0, 1].
(i) Both A and R are true and R is the correct explanatation of A
(ii) Both A and R are true but R is not a correct explanation of A
(iii) A is true but R is false.
(iv) A is false but R is true.
248
(LAS. 1993)
Ans. (i)
UNIT5
Vector Calculus
"This page is Intentionally Left Blank"
Chapter 12
Vector Differential Calculus
Introduction: Let the vector r be a function of a scalar variable t, then
r = f (t)
If only one value of ; corresponds to each value of t, then ~ is defined as a single
valued function of the scalar variable t. If t varies continuously, so does r. As a

result, the end point r describes a continuous curve.
The following illustration make the point clear:
; = a cos t i + a sin t J (Circle)
; = a cos t i + b sin t J (Ellipse)
 A A
r = at
2
i + 2 at j
;=asect i +btant J
(Parabola)
(Hyperbola)
All these are the vector equations of the curves. For different values of t, the end
point of the vector describes the curve as mentioned above.
The vector analysis consists of two parts (i) vector Algebra & (ii) Vector
Calculus. Students have already studied vector algebra, so at present one shall
study only the vector calculus which is very useful while solving problems of
mechanics, fluid mechanics and other branches of Engineering and Technology.
Differentiation of Vector :
Let the vector r be a continuous and single valued function of scalar variable t
(i.e. length of the vector can be determined as soon as a value of t is given) with
o as origin, let the vector r be represented by OA for a certain value of t and let
r + 8 r be represented by OB corresponding to the value t + 8t where 8t is a
  
small increment in t. Then 8t produces an increment (r + 8 r  r) i.e. 8 r in r.
  ar
The increment 8 r is equal to AB. There also the quotient  is a vector if 8 t ~ 0
at
then 8 r ~ 0 and the point B moves towards A to coincide with it and then
chord AB coincides with the tangent at P to the curve.
251
A Textbook of Engineering Mathematics Volume  I
O ~         3 ~       I
dr
dt
or
If the limiting value of the quotient  as o t ~ 0 exists, then this value is defined
ot
 
as the differential coefficient of r with respect to t and the vector r is to be
dr
differentiable and is denoted by . This process is known as differentiation
dt
and the differential coefficient is known as the derivative or the derivative
function.
d;. or. (;+0;);
Thus  = hm
llt
70  = Itm
llt
70 ''
& ~ &
 2
Since dr is itself a vector function of t, its derivative is denoted by d : and is
dt dt

called the second derivative of r with respect to t. Similarly, we can define

higher order derivatives of r .
ILLUSTRATIVE EXAMPLES
 2 2
 A A A dr d r d r
Examplel : If r = a cos t i + a sin t j + t k Find , 2 ' 2
dt dt dt
Solution: we have

dr dAd A d A
=(acost)i+ (asint) J+(t) k
dt dt dt dt
=  a sin t i + a cos t J + k
252
A Textbook of Engineering Mathematics Volume  I
2
dr d ~ d ~ d ~
=(asint) i+(acost) j+(k)
dt
2
dt dt dt
~ 
=  a cos t ia sin t j
2
dr ~ 2 2
Hence 2 = (a cost) + (a sin t) = a
dt

dA   dB   d     
Example 2: If = CxAand = CxB, Prove that (AxB)= Cx(AxB)
dt dt dt

d    dB dA 
Solution: (AxB)=Ax+xB
dt dt dt
     
=Ax(C x B)+(C XA)XB
  
= (A.B)C  (A.C)B + (B.C)A  (B.A)C
.: a x (b x c) = (a.c)b  (a.b)c
  
= (B.C)A  (A.C)B
.,' a.b = b.a
= (C.B)A  (C.A)B
= ex di x A) hence proved.
Example 3 : If ; = (a cos t) i + (a sin t) j + (at tana) k then evaluate
dr x d
2
r and dr d r d r
  [ 2 3]
dt de dt I dt
2
I de
Solution: Given; = (a cos t) i + (a sin t) 1 + (at tana) k
dr ~ ~ ~
:.  = (a sin t)i + (a cos t)j + (a tana)k (i)
dt
2
dr. A A
2 = (a cos t)i + (a sin t)j + (O)k (ii)
dt
3
dr. A
 ~ = (a sin t)i + (a cos t)j (iii)
dr
253
(UPTU 2001)
A Textbook of Engineering Mathematics Volume  I
 2
i k
dr d r
:.x= asin t a cost atana
dt dt
2
acost  asin t 0
= a
2
sin t tana i a
2
cos t tana J + a
2
k
 2
dr d r ~ 2 2 2 2 2 2
:. X2 = (a sinttana) +(a costtana) +(a )
dt dt
= ~ a 4 s i n 2 ttan
2
a + a
4
cos
2
ttan
2
a + a
4
= a
2
.Jtan
2
a+1
= a
2
sec a
a
l
a
2
Also we know [ii, h, cJ = b
l
b
2
c
I
c
2
where ii = al i + a2 J + a3 k
a
3
b
3
c
3
[ , 3]
 asin t a cost
dr d r d r
:. dt' dt
2
'de =
acos t  asin t
asin t  acos t
from (i) (ii) and (iii)
atana
0
0
= a tan a (a
2
cos
2
t + a
2
sin
2
t) expanding the determinant
= a
3
tan a.
Example 4: If r is a unit vector show that
I
" dfl_ldfl rx 
dt dt
df
Solution: We know  is a vector perpendicular to f
dt
" df '"'Idfl' 900. :.rx= r  sm .n
dt dt
(LA.S 1971)
A d A 11 df d A df A f
Where n is a unit vector perpen icular to r as we as  an r , and norm
dt dt
a right handed triad of vectors
254
A Textbook Mathematics Volume  I
A dr Idrl
A
IAI Ab . .
or rx =  n': r =1, r emg umtvector
dt dt
I
A drlldrl IAI Abe'
or r x dt = dt :. n =1, n mg unit vector
Hence proved.
Example 5 : A particle moves along the curve x = 4 cost, Y = sin t, z= 6t Find the
velocity and acceleration at time t = 0 and t = Tt/2
Solution : The position vector of the particle at any time t is given by
;=xi+yj +zk
:.i.e., r=(4cost) i+(4sint) j+6tk
 dr A  A
:.velocity v= =(4sint) i+(4cost) j+6k
dt
and acceleration
2
 d r A 
a= dt
2
=(4cost) i(4sin t) j
From (i) and (ii) we have
at t = 0, = 4 j + 6 k and; =  4 i
att = Tt/ 2. =  4 i + 6 k and; =  4 j
Example 6 : A particle moves along the carve
x = t
3
+1, Y = t2, Z = 2t +5
(i)
(ii)
Where t is the time. Find the components of its velocity and acceleration at time t
= 1 in the direction i + j +3 k.
Solution: Unit vector in the direction of i + j +3 k is
= i + j + 3k _ i + J + 3k
+ (1)2 + (3)2  .Jll

dr d "  A
Now velocity = (xi + Y j + zk)
dt dt
= i.{(e+1)i+t
2
j+(2t+5)k}
dt
= 3t
2
i + 2t J +
= 3i + 2J + 2k at t= 1
(i)
(ii)
H;ence the component of velocity at t =1 in the direction of the vector i + J + 3k is
255
A Textbook of Engineering Mathematics Volume  I
= (3i + 2) + 2k). (Unit vector in the direction of i + J + 3k)
A (i + J + 3k)
= ( 3i + 2j + 2k). 'Jil;l=l='
=3+2+6=Jil
Jil
Again, acceleration
2
d r
dt
2
= :,( ::)
+ 2tJ + 2k)
dt
=6ti +2j
= 6t +2 j at t = 1
Hence, the component of acceleration at t = 1 in the direction of the
vector i + J + 3k is
= (6 i +2 j). (Unit vector along i + ) + 3k)
(
(i + J + 3k)
= 61+2J. c
vll
6+2 8
= =  Answer.
JllJll
Sh th
d r x dr h A r
Example 7 : ow at r x r = 2  were r = 
r
Solution: Since f = 
+
r(  dr)
r
r r
256
A Textbook of Engineering Mathematics Volume 1
Therefore,
. d . (1 d r d )
rx r=rX ; r7 r
r (1 r )
=X dr2"dr
r r r
r x dr r x r
=dr
r2 r3
rxdr
=
2
r
Hence proved.
.rxr=O
Example 8 : A particle P is moving on a circle of radius r with constant angular
dO 
velocity {O =  . Show that the acceleration is {O2 r .
dt
Solution: Let f and j be the unit vectors along two perpendicular radii of the
circle.
If P be any point on the circle such that OP makes an angle 0 with i then the
position vector of P is give by
  
r=OP=OM+MP
= (r cosO) i + (r sinO) j
Where r is the radius of circle and hence constant.
dr dO A dO 
 =  r sinO  i + r cosO  j
dt dt dt
= (r sinoi + r cosO j) {O
dO
As ={O
dt
257
A Textbook of Engineering Mathematics Volume  I
2
 d r de A de 
Hence a = = ( r cose  i  r sine  j) co
dt
2
dt dt
=  (r cose i +r sine j ) co 2
= w
2
r
TICK THE CORRECT ANSWER FROM THE CHOICES GIVEN BELOW

(1) The necessary and sufficient condition for the vector function a (t) to be
da
constant is 
dt


(i) 0 (ii) a

(iii) 2a (iv) 2a
Ans: (i)

    dr
(2) If r = a cos wt + b sin wt, then the value of r x  is
(i) w
2
r
(iii) wa x b
Ans: (iii)
dt
(ii) w
2
r
(iv) wb x a
2
   d r
(3) If r = a cos wt + b sin wt, then the value of 2 is
dt
 
(i) w
2
r (ii) w
2
r
   
(iii) wa x b (iv) wb x a
Ans: (ii)
 
(4) If a is a vector function of some scalar t ~ u c h that a has constant magnitude,

 da
then a.is
dt
(i) 0

(iii) 2 a
Ans: (i)

(ii) a
(iv) 2a

  da
(5) If the direction of a vector function a (t) is constant, then a x  is
dt
 
(i) 0 (ii) a
258
A Textbook of Engineering Mathematics Volume  I

(iii) 2 a (iv) 2a
Ans: (i)
2
    dr
(6) If r = a e
nt
+ b e
nt
, where a, b are constant vectors then 2 is

(i) n
2
r
(iii) 0
Ans: (i)
(ii) n
2
r
(iv) n r
2
 . dr
(7) If r = sin t i + cos t j + t k then 2 is
dt
(i) 0 (ii) 1
(iii) 2 (iv) 3
Ans: (ii)
dt
(8) A particle moves along the curve x=e
t
, y = 2 cos 3t, Z = 2 sin 3t, where t is the
time. The magnitudes of the velocity and acceleration at t = 0 is
(i) I ~ I = &, lal = ~ 3 2 5
(ii) I ~ I = ~ 3 2 5 , lal = &
I
I ~ 7 m
25
(iii) v =  lal= 
325' 37
(iv) I ~ I = /1 lal = rt
'137' 'Ins
Ans: (i)
2
 . dr
(9) If r = (a cos t) i + (a sin t) j + t k then 2 is
dt
(i) 0
(iii) 2a
Ans: (ii)
(ii) a
(iv) 3a
.  .  .. d
(10) If U = t2 i  t j +(2t+ 1) k and V = (2t3) i + j  t k then  (U . V ) when t = lis
dt
(i) 0 (ii) 2
(iii) 4 (iv) 6
259
A Textbook of Engineering Mathematics Volume  I
Ans: (iv)
2
     dA
(11) A = t
m
a +tn b, When a, b are constant vectors, show that, if A and 2 are
parallel vectors then m + n is
(i) 1 (ii) 1
(iii) 2 (iv) 0
Ans : (i)
The differential Operator Del (V)
The operator V is defined as
.0.0 .0
V==i+j+k
ox oy OZ
Vis also known as nabla. It behave as a vector.
SCALAR AND VECTOR POINT FUNCTIONS (UPTU 2001)
dt
A variable quantity whose value at any point in a region of space depends upon
the position of the point, is called a point function. There are two types of point
functions.
(i) Scalar point function: Let R be a region of space at each point of which a
scalar <I> = <I> (x,y,z) is given, then <I> is called a scalar function and R is called a
scalar field.
The temperature distribution in a medium, the distribution of atmospheric
pressure in space are examples of scalar point functions.
(ii) Vector point function: Let R be a region of space at each point of which a
vector V = V (x, y, z) is given them V is called a vector point function and R is
called a vector field. Each vector V of the field is regarded as a localized vector
attached to the corresponding point (x, y, z).
The velocity of a moving fluid at any instant, the gravitational force are examples
of vector point function.
GRADIENT OF A SCALAR POINT FUNCTION:
If <I> (x, y, z) is a scalar point function and continuously differentiable then the
.0<1> .0<1> 0<1>
gradient of <I> is defined as grad <I> = V<I> == i  + j  + k
ox oy oz
= Lio<l>
ox
Thus V<I> is a vector whose rectangular components are
260
ocp ocp and ocp
ox I oy oz
A Textbook Mathematics Volume  I
Geometrical Interpretation of Gradient:
If a surface cp (x, y, z) = c is drawn through any point P such that at each point on
the surface, the function has the same value as at P, then such a surface is called a
level surface through P, for example, if cp (x, y, z) represents potential at the point
(x, y, z), the equipotential surface cp (x, y, z) = c is a level surface.
N
Through any point passes one and only one level surface. Moreover, no two level
surfaces can intersect.
Consider the level surface through P at which the function has value cp and
another level surface through a neighbouring point Q where the value is cp + cScp.
   
Let rand r + 0 r be the position vectors of P and Q respectively, then PQ = or
Now Vcp.o; = (i ocp + J dcp + k
acp
J. (iox + JOY + kOz)
dx dy dz
ocp ocp ocp
= ox+oy+oz
dx dy dz
= ocp
If Q lies on the same level surface as P, then ocp = 0
:.(i) reduces toVcp.or = 0
ThusVcp is perpendicular to every 0 r lying in the surface.
HenceVcp is normal to the surface cp (x, y, z) = C
(i)
LetVcp = IVcplN, Where N is a unit vector normal to the surface. Let PA=on be
the perpendicular distance between the two level surfaces through P and Q.
Then the rate of change of cp in the direction of normal to the surface through P is
ocp. oCP. Vcp.or .
 = hmon+o  = hm1\n+o by (1)
dn on on
. IVcplN.orl I I
= hm = Vcp
1\n+O on
261
A Textbook of Engineerins Mathematics Volume  I
.,' N ,of = INlloflcose
=Ioflcose
= on
,'IV<p1 =:
Hence the gradient of a scalar field <P is a vector normal to the surface <P = c and
has a magnitude equal to the rate of change of <p along this normal.
Directional Derivative:
Let <p (x, y, z) be a scalar point function and s represent a distance from any point
P(x, y, z) in the direction of a unit vector a, then d<p is called the directional
ds
derivative of q, in the direction of a.
The directional derivative of a scalar point function is a scalar and of a vector
point function is a vector.
Theorem : The directional derivative of a scalar field <p at a point P(x, y, z) in the
direction of a unit vector a is given by
d<p ' d
 = a.gra <p
ds
= a .v<p
Proof: Since a is a unit vector at the point P(x, y, z), therefore
, zdx ~ d y k' dz
a=l+J+ 
. ds ds ds
Where s represents a distance from P in the direction of a .
Now a. grad <p = a,V<p
= (i dx + J dy + k dz).(i dq, + r
1
<P + k d<P)
ds ds ds dX ay dZ
= dx d<p + dy d<p + dz d<p
ds . dX ds' dy ds . dZ
= d<p Hence Proved.
ds
Example 1: If r = xi + y} + zk show that
I i
(i) Vr = r (ii) V =
r r2

(iii) Vrn = nr
n

2
r
262
(UPTU 2007)
A Textbook of Engineering Mathematics Volume  I
Solution :  Since; = x i + Y J + Z k
1;1 = + y2 + Z2
I 2 2 2
="x +y +z
=:> r2 = x2+ y2 +Z2
(
0 0 0 )
(i) Vr = i  + j  + k  r
ox ay az
.ar .or or
=i+j+k
ax oy az
.x .y . z
= i+ j+ k
r r r
xf+Yl+
zk
=
r
r
=
r
=r
(ll)V= 1+ J+k 
.. 1 . a a ) 1
I ax ay oz r
=i;x(;)+j:y(;)+k:z(;)
.lar .lar lar
= 1 Jk
r2 ox r2 oy r2 oz
= j y+
r2 r r r
=
xf+yj+zk
r
= r3
r
= ?"
.. : r2 = x
2
+ y2 +Z2
ar
2r= 2x
ax
ar x
ax r
.. ar y
snmlarly  = 
oy r
ar z
and =
oz r
263
A Textbook Mathematics Volume  I
r
','r=
r
n a A a J n
(iii) Vr = i+ j+ k r
ax ay az
n1 ar n1 ar A n1 ar
=lnr +Jnr +knr 
ax ay az
n1 X n1 Y k" n1 Z
=lnr +Jnr + nr 
r r r
= i nr
n

2
x + 1 nr
n

2
y + k nr
n

2
z
= nr
n

2
(xi + YI + zk)
n2
= nr r
Example 2: If <I> (x, y, z) = 3x2y  y3z2, find V<I> at the point (1, 2, 1)
(
" a A a "a J( 2 3 2)
Solution: V<I>= i+j+k 3x yy z
ax ay az
" a (2 3 2) A a (2 3 2) "a (2 3 2)
= i  3x Y  y' z + j  3x Y  Y z + k  3x Y  Y z
ax ay az
" "( 2 2 2) A( 3)
= i (6xy) + j 3x  3y z + k 2y z
:.At (1,2, 1) we have
Vcj> = i [6 (1) (2)] + j [3(1)23(2)2 (1)2] + k [2 (2)3 (1)]
=  12 i 9 j 16 k Answer
 
Example 3 : Show that V(a.r) = a where a is a constant vector.
Proof: Let a = a1 i +a2 J +a3k
Then a.r = (a1 i +a2J +a3k). (xi + YJ +zk)
= a1X +a2Y+ a3Z
, ( ) (" a A a A a J ( )
Therefore V a.r = i+j+k a\x+a
2
y+a
3
z
ax ay az
= i a1 + J a2 + k a3
= a, hence proved.
   
Example 4: If ,a and b be constant vectors, then show that grad [r a b]= a x b
Solution : Let ; = xi +y J +z k, a = a1 i +a2J +a3 k
264
A Textbook of Engineering Mathematics Volume  I
and b = bl i + b2 J + b3 k
x y z
Then <I> = [;; bJ = a
l
a
2
a
3
b
l
b
2
b
3
= x(a2b3  a3b2) + y(a3bl  aIb3) + z(aIb2  a2bl)
Therefore grad <I> = grad [;; bJ = i 0<1> + J 0<1> + k 0<1>
ox oy oz
= i (a2b3  a3b2) + J (a3bl  aIb3) + k (aIb2  a2bl)
0<1>
.:  = a2b3  a3b2
ox
0<1>
 = a3bI  aIb3
oy
0<1>
and  = aIb2  a2bI
OZ
= a x b Hence proved
Example 5: If <I> (x, y) = log ~ X 2 + y2 show that
   
r  (k.r)k
grad <I> = { ____ } { ____ }
r(k.r)k . r(k.r)k
   
Solution:  we have r = x i +y j +z k (i)
Therefore r. k = z (ii)
1
Now <I> = log (x
2
+ y2)
2
0<1> 1 x
.:  = 2 2 .2x = 2 2
ox 2(x + Y ) x + Y
0<1> y 0<1>
similarly  = 2 2'= 0
oy x + Y OZ
0<1> 0<1>  o<j>
Thus grad <j> = i+ j+ k
ox oy oz
x  x  
= 2 2i+ 2 2j+Ok
x +y x +y
265
A Textbook of Engineering Mathematics Volume  I
 
_ xi + yj
 x2 + y2
 
rzk
 
rzk
= (_ _) (_ _) by
rzk . rzk
(i)
Now by replacing z by r. k we get
;(k.;)k
grad <I> = {_ ( __ )_} {_ ( __ )_}
r k.r k . r k.r k
Hence proved.
Example 6: Find the directional derivative of <I> = xy + yz +zx in the direction of
vector i +2) +2 kat (1, 2, 0)
Solution: Since we know
directional derivative = a. grad <I>
A d<l> A d<l> A d<l>
Now grad <I> = i+ j+ k
dX dy dZ
= (y + z) i + (z + x) J + (x + y) k
= 2 i + J + 3 k at (1, 2, 0)
A i+i+2k
Also a = _,1 
3
:. directional derivative =! (i +2 J + 2 k ).(2 i + J + 3 k )
3
1
= (2+2+6)
2
10
=Answer.
3
Example 7: Find the directional derivative of the function <I> = x
2
 y2 + 2z2 at the
point P (1, 2, 3) in the direction of the line PQ, where Q is the point (5, 0,4).
(UPTU 2000).
Solution : The position vectors of the points P and Q are respectively
i + 2 J + 3 k and 5 i +4 k
PQ = (5 i +4 k )  (i + 2 J + 3 k )
266
A Textbook of Engineering Mathematics Volume  I
= 4t 2J +k
The unit vector a along PQ is given by
4i  ij + k 4i 2) + k
a= ='
+ (_2)2 + (1)2 J2i
acj> acj> A acj>
Now grad cj> = i+ j+ k
ax ay az
=2xt2Yl+4zk
directional derivative = a . grad cj>
4i  2) + k . . .
= . (2x i  2y j + 4z k )
.fii
= Sx+4y+4z
J21
= S.l + 4.2 + 4.3 at (1,2,3)
J21
Example 8: Find the directional derivative of cj> = xL 2y2 + 4Z2 at (I, I, 1) in the
direction 2 i + J  k . In what direction is the directional derivative from the point
(1,1, 1) is maximum and what is its value?
Solution: We have, directional derivative in the direction of a is a. grad cj>
acj> A acj> A acj>
Hence grad cj> = i+ j+ k
ax ay az
= 2x i  4y } + Sz k
=2i 4} 8kat(l,l,1)
2t+jk
and .1= 16
Therefore, directional derivative
=( 2i 1
k
}(21 4) 8k)
44+8 8
= J6 =J6
267
A Textbook of Engineering Mathematics Volume  I
Again, directional derivative is maximum along the normal i.e. along grad <I> i.e.
2 i 4 J  8 k and hence the maximum value of directional derivative is
/grad<l>/ = .j4+16+64 = 2m Answer.
Example 9 : Find the values of the constants a, b, c so that the directional
derivative of <I> = axy2 + byz + CZ
2
X
3
at (1, 2, 1) has a maximum magnitude 64 in a
direction parallel to zaxis.
(lAS 2002, 2006)
Solution : We know that the directional derivative is maximum along the
normal i.e. along grad <I>
here, we have
grad <I> = (ay2 + 3CZ
2
X2) i +(2 axy + bz) J + (by + 2czx
3
) k
= (4a +3c) i +(4a  b) J +(2b  2c) k at (1, 2, 1)
But directional derivative is maximum along zaxis. Hence the coefficients of
i and J should be zero.
:. 4a + 3c= 0 and 4a  b = 0
:.grad <I> = (2b  2c) k
Also maximum value of directional derivative = /grad<l>/
:. 64 = 2(b  c)
=> b  c = 32
Solving these three equations, we get
a= 6, b = 24, c = 8 Answer.
Problem 10 : Find the angle between the normals to the surfaces
x
2
+ y2 + Z2 = 9 and z = x
2
+ y2  3 at the point (2, 1, 2)
Solution : Let the given surfaces represented by
<1>1 == x
2
+ yL z3 and <1>2 == x
2
+ y2+ z29
 : a ~ ~ a ~  a ~
:.n1 = V <1>, =l+J+k
ax ay az

=2xj +2y jk
= 4 i 2 j  k at the point (2, 1, 2) similarly, we have
   
n2 = V<I>2 = 4 i 2 j +4k at the point (2, 1, 2)
(D.P.T.U 2002, LA.S 1973)
   
If n1 and n2 be the normals to the surfaces <1>1 and <1>2, then nl = V<I>l and n2 = V<I>2
Let e be the angle between the normals to the surface at the given point then
~ 1 . ~ 2 = 1 ~ 1 1 1 ~ 2 1 cos e
268
A Textbook ofEllgineering Mathematics Volume  I
n1.n2
:. cos e = 1;11\;21
_ (4i2Jk) (4i2J+4k)
 .J16+4+1 .J16+4+16
16 + 4 4
61
8
 31
e = cos
1
( 3 ~ ) Answer.
Problem 11 : Find the directional derivative of a function <!> == x
2
y3 Z4 at the
point (2, 3, 1) in the direction making equal angles with the positive x, y, & Z
axis.
Solution : Given <!> = x
2
y3z4
"o<!> "o<!> "o<!>
Now grad <!>=i+j+k
ox oy oz
= 2
xy
3Z4 i +3X
2
y
2
z
4 J +4x
2
y3z3 k
If a be the unit vector in the required direction and a be the angle which a
makes with the axes, then
a = (cos a) i + (cos a) J +(cos a) k
where cos
2
a + cos
2
a + cos
2
a = 1
1
which gives cos a = J3
1 (" " ")
:. a = J3 i + j + k
:.directional derivative = a. grad <!>
1 " " " A " "
= J3 (i + j + k ).(2xy2z4 i +3X
2
y2z4 j +4x
2
y3z3 k )
1
=  (2xy3Z4+3x
2
y2Z4+4x
2
y3Z3)
J3
= ~ (108 +108  432) at the point (2, 3, 1)
216
=  J3 Answer.
269
A Textbook ofEllzineerinz Mathematics Volume  I
Example 12 : Find a unit vector which is perpendicular to the surface of the
paraboloid of revolution.
z = X2+y2 at the point (I, 2, 5)
(B.P.S.C 1997)
Solution : 4> = x2+y2_z
A 04> A 04> A 04>
:. grad 4> = i  + j  + k 
ox oy oz
= i.2x + J .2y  k.l
= 2 i +4 J k at point (1, 2, 5)
Hence unit normal = ,grad4>,
grad4>
_ 2i + 4J k
 "'4+16+1
2i + 4j  k
= Answer.
J21
Problem 13 : What is the greatest rate of increase of 4> == xyz2 at the point (I, 0, 3).
A 04> A 04> A 04>
Solution: grad 4> = i+ j+ k
oX oy oz
= i yz2 + J XZ2 + k 2xyz
= i 0 + J 9 + k 0 at (I, 0, 3)
= 9 j
Since we know the greatest rate of increase of 4> = 1V'4>1
= J(9)2
=9Answer.
Divergence of a vector :
If V (x, y, z) is any continuously differentiable vector point function, then the
divergence of V , written as div V or V'. V is defined by
  (A 0 A 0 A 0)
divY=V'Y= i+j+k.Y
ox oy oz
270
A Textbook of Engineering Mathematics Volume 1
av av av
=i+j+k
ax ay az
av
ax
v. V is a scalar product of the operator Vwith the vector V.
Physical Interpretation of Divergence
Consider a fluid having density p = p (x, y, z, t) and velocity v = v (x, y, z, t) at a
point (x,y,z) at time t. Let V = p v, then V is a vector having the same direction
as v and magnitude plvl . It is known as flux. Its direction gives the direction of
the fluid flow, and its magnitude gives the mass of the fluid crossing per unit
time a unit area placed perpendicular to the direction of fJow.
z
Ox
X
Consider the motion of the fluid having velocity V = V xi + V y J + V z k at point
P(x, y, z). Consider a small parallelopiped with edges Ox, oy, oz parallel to the
axes with one of its corners at P. The mass of the fluid entering through the face
Fl per unit time is Vy Ox oz and that flowing out through the opposite face F2 is
(
avv )
Vy + Ily Ox oz = Vy + ay oy Ox oz by using Taylor's Theorem.
:. The net decrease in the mass of fluid flowing across these two faces
(
av, )
= Vv + " oy Ox oz  Vy Ox Oz
" ay
271
A Textbook of Enzineering Mathematics Volume  I
aV"
= 8x8y8z
ay
similarly, considering the other two pairs of faces, we get the total decrease in the
ffl
d d th 11 1 . d .. (aVx ~ a V y k
A
avz)s: s: s:
mass 0 Ul mSI e e para e opIpe per umt time l ax + J ay + az uxuyuz.
Dividing this by the volume 8x 8y 8z of the parallelopiped, we have the rate of
1 f fl
d .. aV
x
aV. aV
z
osso Ul perumttime =+' +
ax Oy az
=divV
Hence div V gives the rate of outflow per unit volume at a point of the fluid.
If the fluid is incompressible, there can be no gain or loss in the volume element.
Hence div V = 0 and V is called a solenoidal vector function. Which in known in
Hydrodynamics as the equation of continuity for incompressible fluids.
Note: Vectors having zero divergence are called solenoidal and are useful in
various branches of physics and Engineering.
(U.P.T.U 2002, 2003, 2006).
CURL OF VECTOR POINT FUNCTION
The curl (or rotation) of a differentiable vector point function V is denoted by
curl V and is defined as
_ _ (A a A a A a) 
curl V = V' x V = i  + j  + k  x v
ax ay az
~ av ~ av A aV
=IX+Jx+kx
ax ay az
~ aV
=LIX
ax
The curl of a vector point function is a vector quantity if V = VI i + V 2 J + V3 k
Then
272
A Textbook ofEllgineering Mathematics Volume  I
j
k
a a a
=

ax ay az
V
1
V
2
V3
= i (
aV
3 _ aV
2
) + J (aVl _ aV
3
) + k (aV2 _ aV
1
)
ay az az ax ax ay
PHYSICAL INTERPRETATION OF CURL: Consider a rigid body rotating
about a given axis through 0 with u n i ~ o r m angular velocity 0).
w
o
Let (;;=0)1i+0)2J +0)3k
The linear velocity V of any point P(x, y, z) on the rigid body is given by
V=O)xr
Where r = i x+ J y+ k z is the position vector of P
,'. V = 0) x r
k
x y z
.,' curl V = curl (00 x r) = yo x (00 x r)
273
A Textbook ofEllgineering Mathematics Volume  I
=
a
ax
j
a
ay
k
a
az
(02
Z
 (03Y (03
X
 (OI
Z
(OIY  (02
X
= ((01 + (01)i + ((02 + (02)J + ((03 + (03)k
= 2( (Ol i + (021 + (03k )
:. (01,0>2, W3 are constants
= 2 (0
 1 
:. (0 =  curl V
2
Thus the angular velocity at any points is equal to half the curl of linear velocity
at that point of the body.
(U.P.T.U 2001).
Note : If curl V = 0, then V is said to be an irrotational vector, otherwise
rotational. Also curl of a vector signifies rotation.
VECfOR IDENTITIES
(1) grad (ii.b) = (ii.V)b+(b.V)ii+ii x curl b+ b x curl;
where ii and b are the differentiable vector functions
(I.A.S 2004, U.P.P.C.S 2004)
Proof:
(
 ) A a ( ) A a ( _.) A a ( )
grad a.b = i a.b + j; a.b +k a.b
ax uy az
A aa   ab
(
 )
=Li ax'b+a'ax
A(_ aii) A(_ ab)
=Li b ax +Li a ax
(i)
.  ( ~ ab) ( a b ) ~ ( ~ ) ab
Agam ax IX = a. 1 a.1 
ax ax ax
or a I=ax IX + a.1 
(
 a b ) ~  ( ~ ab) (  ~ ) ab
ax ax ax
274
A Textbook of Engineering Mathematics Volume  I
(
 ob )  ( ob ) (_. 0 )
:.L a
ox
i=ax Lix
ox
+L a.i
ox
b
=; x curl b+(a.V') b (ii)
similarly
r( b :)i = b x curl ; +( b S') ; ......... (iii)
Hence from (i) , (ii) & (iii) we get
grad ( a. b) = (a.V') b + ( b.V' ) ci + a x curl b + b x curl;
Hence proved.

(2) If a is a vector function and u is a scalar function then
  
div (u a) = u diva + (grad u) . a
(U.P.T.U 2004, B.P.S.C 1995)
Proof :
Let; =a
1
i +a2) +a3k
:. +ua2j +ua3k)
i + ua
2
J +ua
3
k)
ox oy oz
000
= (ua
1
)+(ua
2
)+ ;(ua
3
)
ox oy oZ
OU oa
l
OU oa
2
OU oa
3
=a
l
+u+a
2
+u +a
3
+u
ox ox oy oy 3 OZ OZ
= (ou a + OU a + OU a )+ u(oa
1
+ oa
2
+ oa
3
)
ox 1 oy 2 OZ 3 ox oy OZ
= (ou i + ou ) + ou k). (a
1
i + a
2
1 + a
3
k)+ u(i + Ji.. + k (a
1
i + a
2
l + a
3
k)
= V'uci +uV'. a
 
=uV'a+V'ua
  
Thus div (u a) = udiv a + (grad u) a
Hence proved.
 
(3) Prove that div (a x b) = b. curl a  a. curl b
[U.P.T.U 2003, B.P.S.C 1993).
Proof : div ( a x b ) = V' .( a x b)
275
A Textbook of Engineering Mathematics Volume  I
a  
=Ll'ax(axb)
= Ll. xb+ax
[aii   ab]
ax ax
L
~ aii ~  ab
= J.x b+ Ll.aX
ax ax
aii  . ab
= L 1 X ax .b  L 1 X ax' a
   
:a.bx c= axc.b
aa  aa
(
) ( )
= Li x ax .b Li x ax .a
= curl a . b  curl b. a
Thus div (a x b) = b.curla  a.curl b
Hence proved.
(4) If a is a vector function and u is a scalar function then
  
Curl (u a) = u curl a +(grad u)x a
Proof: Let; = at i +a2J +Cl3k
:. V.(u;) = V. ( u at i + U a2J + u a3 k )
(
a a a J ( . . .)
= i  + j  + k  x ua
t
i + ua
2
j + ua
3
k
ax ay az
j k
a a a
=

ax ay az
ua
t
ua
2
ua
3
276
[V.P.T.V. (CO) 2003]
A Textbook of Engineering Mathematics Volume  I
j
k k
d d d dU dU dU
=U  
+
dX dy dZ dX dy dZ
a
1
a
2
a
3
a
1
a
2
a
3
 
= u (Vx a) + (Vu) x a
  
Thus curl (u a)= u curl a +(grad u) x a
Hence proved.
(5) Prove that Curl (a x b) = (b.V)a (a.V)b+ a div b b div ~
(IA.S 2000, u.P.P.CS 1996)
Proof: Curl (a x b)= i x ~ ( a x b)+ J x ~ ( a x b)+ k x ~ ( a x b)
dX dy dZ
=}:iX xb+ax
A [da   db]
dX dX
A (da ) A ( db)
=}:i x dX x b +}:i x a x dX
=}: (Lb) L b+ L a(La)
[
A_ da (A da) (A db) A_ db]
dX dX dX dX
.: a x (b x c) = (a.c)b  (a.b)c
(
_A d )   A da  A db (  Ad)
= }:b.i dX ab}:i
dX
+a}:L
dX
 }:a.i dX b
    
= (b.V') a  b diva + a div b  ( a . V) b
  
= (b . V) a  ( a . V) b + a div b  b diva
     
Thus curl ( a x b ) = (b . V) a  ( a . V) b + a div b  b diva
Hence proved.
277
A Textbook of Engineering Mathematics Volume  I
SECOND ORDER DIFFERENTIAL OPERATORS, LAPLACE'S OPERATOR
V2 :
a
2
0
2
0
2
The operator V
2
=  +  +  is called the Laplace's operator and the
ox
2
ay2 OZ2
equation 0 is called the Laplace's equation. Now we shall prove some
results of second order differential operators.
(1) div grad q, =v.v q, =
Proof : We have
" oq, " oq, "oq,
Vq, = i+ j+ k
ox oy oz
(
" 0 "0 "0 J (" oq, " oq, "oq,J
.: V . Vq, = i ox + j ay + k OZ . i ox + j ay + k OZ
02q, iq, iq,
=++
ox
2
ay2 OZ2
= V'2q,
Thus div grad q, =V2q" hence proved.
(2) Curl grad q, = V'x(V'cp) = 0
Proof: Curl grad q, = V'x(Vcp)
a a a )
V'<I> = i ax + j ax + k ax (3x+2y+z6)
= 3i +2 J + k
3i + 2j + k
Unit normal vector n = J
(3)2 + (2)2 + (1)2
3i +2J + k
= J14;=:14=
:.JfCurlF.nds= Jf(2i+k).
...... 14
where S is the triangle ABC
= ( 6 + 1) Jf ds
...... 14 5
7
= r:;; (Area of !!. ABC)
...... 14
It is difficult to find area of!!. ABC, so we change ds to d
l
,where Sf dxdy is
n.k
area of!!. OAB
345
7 Ifdxdy
= M s In.kl
7 If dxdy
= M 11M
=7If dxdy
R
A Textbook of Engineering Mathematics Volume  I
= 7 (Area of ~ OAB)
1
= 7 ( x2
x
3)
2
=21
Since we know Stoke's theorem is given by
pF.dr = If CurlF.nds
Therefore, from (i) and (ii) we have
p[(x + y)dx +(2x z)dy +(y + z)dzJ = 21
c
(i)
(ii)
Example 8: IfF = (xz) i +(x3+yZ) J 3xy2k and S is the surface of the curve
If
3na4
z = a  ~ X 2 + y2 above the xyplane, show that Curl P.ds =
s 4
Solution: Here P = (xz) i +(x
3
+yz) J 3xy2k (i)
By Stoke's theorem, we have
If CurlF.ds = pF.dr
where S is the surface x2+y2 = a
2
, z=o above the xyplane.
F.dr = [(xz) i +(x
3
+yz) J  3xy2 k ].(dx i +dy J +dz k)
(ii)
= (xz)dx + (x
3
+yz)dy3xy2 dz (iii)
Let x = a cosS so that dx = a sinS dS and y = a sinS so that dy = a cosS dS
211
pP.dr= f {(acosS)(asinSdS)+a
3
cos
3
S. a cos SdS}
c 0
.,' z = 0 ,'. dz = 0
346
Vector Integral Calculus
2n 2n
= _a
2
J sinS cosSdS + a
4
J cos
4
SdS
o 0
2n
= a
4
J cos
4
SdS
o
nl2
= 4a
4
J cos
4
SdS
o
=4a
4
~ ~
2[3
[131[1
=4a4122212
4
4 3
=a 1t
4
3 4
=1ta
4
~ JJ CurIF.ds = ~ 1 t a 4
s
Using (ii) & (iv)
EXERCISE
(iv)
1. Suppose F = x
3
i +y J +z k is the force field. Find the work done by F along the
line from the (1, 2, 3) to (3, 5, 7).
(U.P.T.U.2005)
Answer. 50.5 units
2. Evaluate J F.dr where F = xy i +(X2+y2) J and C is the arc of the curve y = x24
c
from (2, 0) to (4,12) in the xyplane.
Ans.732
3. Evaluate J F.dr for F = 3x
2
i +(2xzy) J +z k along the path of the carve x
2
= 4y, 3x
3
c
= 8z from x = 0 to x = 2
Ans.16
J
   iy  jx
4. Compute F.dr, where F = 2 2 and C is the circle X2+y2=1 traversed
r x +y
counter clockwise.
Ans. 21t
347
A Textbook Mathematics Volume  I
5. Find the circulation ofF round the curve C, where F = y i +z J +x k and C is the
circle X2+y2 = 1, Z = 0
Ans. n
Hint Circulation = f F.dr
6. Find the work done in moving a particle once around a circle C in the xy
plane, if the circle has centre at the origin and radius 2 and if the force field is
given by F = (2xy+2z) i +(X+yZ2) J +(3x2y5z) k
Ans.81t
7. Evaluate ff (yzi + zXJ + xyk ).ds where S is the surface of the sphere X2+y2+z2 =a
2
in the first octant.
3a
4
Ans.
8
8. ifF = (X2_y2) i +2xy J and;: = xi +y J find the value of IF. df around the
rectangle boundary x = 0, y = 0 x = a, y = b
(U.P.T.U 2002)
Ans.2ab
2
9. Evaluate f[(cosxsiny  xy)dx +sinxcosydy JbY Green's theorem where C is
c
the circle X2+y2 =1
Ans.O
10. Evaluate IF. it ds where F = 4xy i + yz)  xZk where and S is the surface of the
5
cube bounded by the planes x =0, x = 2, Y = 0, Y = 2, z = 0, z = 2
Ans.32
11. Prove that Iff CurlF dv = ff it x F ds
12. Prove that I it x (a x r)ds = 2aV , where V is the volume enclosed by the surface
s
S and a is a constant vector.
13. Verify Gauss's divergence theorem for F = y i +x J +Z2 k and S is the surface of
the cylinder bounded by X2+y2 = a
2
; z = 0; z = h
14. Verify divergence theorem for F = 4xz i _y2 J + yz k and S the surface of the
cube bounded by the planes x=O, x = 2, Y = 0, Y = 2, z = 0, Z = 2
348
Vector Integral Calculus
15. Verify Gauss's theorem and show that
5
H(X
3
 yz)i  2X2Yl + 2k ].nds =; where S denotes the surface of the cube
5
bounded by the planes x = 0, x = a, y = 0, y = a, z = 0, z = a.
16. Evaluate Sf (yzi + zXJ + xy k ) ds where S is the surface of the sphere x2+y2+Z2 =
s
a
2
in the first octant.
(U.P.T.U 2004)
Ans.O
17. Verify the divergence theorem for the function F = 2x2y i y2 J + 4XZ2 k taken
over the region in the first octant bounded by y2+Z2 = 9 and x = 2.
18. Evaluate by stoke's theorem p (ex dx + 2y dy  dZ) where C is the carve
X2+y2 = 4, Z = 2
Ans.O
19. Apply stoke's theorem to find the value of f (ydx + zdy + xdz) where C is the
curve of intersection X2+y2+Z2 = a
2
and x+z = a
(JNTU 1999)
,
na
Ans. J2
20 Evaluate p 2y
3
dx + x
3
dy + zdz where C is the trace of the cone intersected by
<'
the plane z = 4 and S is the surface of the cone z = Jx
2
+ y2 below z = 4
(PTU 2006)
Ans.192n
21. Use Stoke's theorem to evaluate f Curl'F.nds over the open hemispherical
surface X2+y2+Z2 =a
2
, z > 0 where F = y i + ZX J + y k
Hint: The boundary of hemispherical surface is circle of radius a in the plane z =
O. The parametric equations of the circle are x = a cosS, y = b sinS, z = 0
Here r = xi + y J + Z k ,d r = dx i + dy J + dz k , X2+y2 = a
2
, z = 0
using stoke's theorem
fF.dr = f (yi + zx) + yk).(dxi + dy} + dzk)
c ('
we get
~ f CurlF.nds = f ydx
5 c
349
A Textbook of Engineering Mathematics Volume  I
271
= f asin(asinS)dS
o
2 271
=  ~ f 2sin2 SdS
2 0
2 271
=  ~ f(1cos2S)dS
2 0
=  ~ [ S _ sin2SJ
271
2 2 0
= 7m
2
Answer
Tick the Correct Answer of the Choices Given Below :
1. The value of f F.dr for F = 3x
2
i + (2xzy) J + zk along the path of the curve x
2
=
c
4y, 3x
3
= 8z from x = 0 to x = 2 is
(i) 8 (ii) 16
(iii) 4 (iv) 10
Ans. (ii)
2. IfF = i;'  j ~ evaluate fF.dr where C is the circle X2+y2 =1 transversed counter
x +y c
clockwise
(i) 1t
(iii) 2 1t
Ans. (iv)
(ii) 1t
(iv) 2 1t
3. The value of f F.dr for F = x
2
i +xy} and C is the curve y2 = x joining (0, 0) to
(1, 1) is
(i) 2.
12
(
... ) 5
111 
12
Ans. (i)
c
(
.. ) 3
11 
12
(iv) 1
4. If F = (3X2+6y) i 14yz J + 20xz
2
k , Then the value of line integral f F.dr form (0,
c
0,0) to (1, 1, 1) along the path x = t, Y = t2, Z = t
3
is
(i) 2 (ii) 5
(iii) 7 (iv) ~
2
Ans. (ii)
350
Vector Intewal Calculus
5. The value of the line integral J[(x
2
+ xY)dx +(X2 +y2)dYJ ,where C is the
c
square formed by the lines y = 1 and x = 1 is
(i) (ii) 1
(iii) 1 (iv) .!
2
Ans. (i)
6. The circulation ofF round the curve C, where F = y i + J + x k and C is the
circle X2+y2 =1, z =
(i) 1t (ii) 1t
(
. ) 1t (.) 1t
III  IV 
2 2
Ans. (ii)
7. IfF = 2y i  Z J + x k then the value of J F x dr along the curve x= cos t, Y = sin t,
Z =2cost from t = t =1t/2 is
(i) i + J (ii) i  J
c
(
... ) (2 (
III 4' 1+ 1t"2 J (iv)
Ans. (iii)
8. The value of J F.dr where F = xy i +(X2+y2) J and C is the x axis is from x = 2 to x
c
= 4 and the line x =4 from y =0 to Y =12 is
(i) 768 (ii) 785
(iii) 763 (iv) 764
Ans. (i)
9. The circulation ofF around C, where F = ex sin y i +e
X
cos y J and C is the
rectangle whose vertices are (0,0), (1, 0), (1, ), (0, ) is
2 2
(i) (ii) 1
(iii) 1 (iv) .!
2
Ans. (i)
10. The vector function A defined by A =(sin y+z cos x) i +(x cos y+sinz) J + (y
cos z+sin x) k is irrotatinal. Then function cp is given by (Given A = Vcp)
(i) x sin y  y sin z  z sin x + c
(ii) x sin y + y sin z + z sin x + c
(iii) x cos Y + Y cos z + z sin x + c
351
A Textbook of Engineering Mathematics Volume  I
(iv) x cos Y + Y cos z  z sin x + c
Ans. (ii)
11. Consider a vector field F = (X2  y2 + x) i (2xy+y) J Then the field is
(i) Solenoidal
(ii) irrotational
(iii) Solenoidal and irrotational
(iv) None of the above
Ans. (iii)
(V.P.T.V.2009)
12. Consider a vector field F = (X2  y2 + x) i (2xy+y) J . Then cjl is given by:
(GivenF = V<I
x
3
x2 y2
(i) xy2 +
3 2 2
3 2 2
(
") X 2 X Y
11 +xy 
3 2 2
3 2 2
(
"') X 2 X Y
III xy 
3 2 2
. x
3
2 x
2
y2
(lV) xy 
3 2 2
Ans. (i)
13. The value of ffF.nds ,whereF = z2i +xy J y2k and S is the surface bounded
s
by the region x2+y2 = 4 z = 0, z = 3 is
(i) 26
(iii) 13
Ans. (ii)
(ii) 26
(iv) 13
14. IfF = 4xz i  y2 J + yz k and S is the surface of the cube bounded by the planes
x = 0, x = 1, Y = 0, Y =1, z=O, z=1 Then the value of ffF.nds is
5
(
") 3
11 
2
(i) !
2
(iv) Z.
2
(
"') 5
III 
2
Ans. (ii)
15. If V is the volume enclosed by a closed surface S and F = xi +2y J +3z k . Then
fF.nds is
5
(i) 3V
(iii) 9V
Ans. (ii)
(ii) 6V
(iv) 12V
352
Vector Integral Calculus
16. The value of F.nds ,where S is a closed surface is
(i) 3V
(iii) 9V
Ans. (i)
5
(ii) 6V
(iv) 0
17. The value of J nds ,where S is a closed surface is
(i) 3V
(iii) 9V
Ans. (iv)
5
(ii) 6V
(iv) 0
18. The value of J n .(v x F), where F is a vector point function and S is a closed
surface is
(i) 3V
(iii) 9V
Ans. (iv)
s
(ii) 6V
(iv) 0
19. The value of J F.n ds , for F = 4x i  2y2 J +Z2 k taken over the region S bounded
5
by X2+y2 = 4, Z =0 and Z =3 is
(i) 38n
(iii) 83n
Ans. (ii)
(ii) 84n
(iv) 48n
20. The value of F x n ds , for any closed surface S is
(i)1
(iii) 3
Ans. (iv)
s
(ii) 1
(iv) 0
21. The value of Jnx VcI>ds ,for a closed surface S is
(i) 2a
(iii) 3 a
Ans. (iv)
5
(ii) 2a
(iv) 0
22. The value of J n x (a x r)ds where V is the volume enclosed by the surface S
5
and a is a constant vector is
(i) 2a
(iii) 2 a
(ii) 2a V
(iv) 0
353
A Textbook of Engineering Mathematics Volume  I
Ans. (ii)
23. The value of J( e
x
sin ydx + e
x
cos ydy) where C is the rectangle with vertices
c
(0, 0), (1t, 0), (1t, ~ ), (0, ~ ) is
2 2'
(i) 2(e
1t
1) (ii) 2 (e
1t
l)
(iii) 2(e
1t
+1) (iv) 2(e" +1)
Ans. (ii)
24. The value of f (X2y dx + x
2
dy) ,where C is the boundary described counter
c
clockwise of the triangle with vertices (0, 0), (1, 0), (1, 1) is
(i) 0 (ii) !
2
(
... ) 5
111 
12
Ans. (iii)
(iv) !
2
25. The value of f (X2 + xy)dx + (X2 + y2 )dy where C is the square formed by the
c
lines y = 1, x = 1 is
(i) 0 (ii) !
2
(
... ) 5
111 
12
Ans. (i)
(iv) !
2
26. The value of f (ex dx + 2y dy  dz) , where C is the curve X2+y2 = 4, z = 0 is
c
(i) 0 (
.. ) 1
11 
2
(
... ) 5
111 
12
Ans. (i)
27. The value of the f (xydx + xy
2
dy) where C is the square in the xyplane with
c
vertices (1,1) (1, 1),'(1, 1), (1, 1) is
(
.) 1 ( .. ) 2
1  11 
3 3
(
... ) 4 (. ) 4
111  IV 
5 3
Ans. (iv)
354
Vector Integral Calculus
28. The value of f P.df where P = y2 i + X2]  (x+z) k and C is the boundary of the
5
triangle with vertices at (0, 0, 0), (1, 0, 0), (1, 1, 0) is
(
.) 1 (") 2
I  11 
3 3
(
"') 4 (. ) 4
111  IV 
5 3
Ans. (i)
29. The value of f (yz dx + xz dy + xy dz) , where C is the curve x
2
+ y2 =1, z = y2
(i) a
(iii) 2
Ans. (i)
c
(ii) 1
(iv) 2
30. The value o f ~ r . d r is
(i) 0
(iii) 2
Ans. (i)
c
(ii) 1
(iv) 3
31. A necessary and sufficient condition that line integral fA.dr = 0 for every
closed curve C is that
(i) div A = 0
(iii) div A * 0
Ans. (ii)
(ii) curl A = 0
(iv) curl A * 0
c
32. The value of surface integral If (yzdydz + zxdzdx + xydxdy), where 5 is the
5
surface of the sphere x2+y2+z2=1 is
(i) 41t (ii) 0
3
(iii) 41t (iv) 121t
Ans. (ii)
33. If P = axi + by] +czk a, b, c are constants, then If F.ds where 5 is the surface of
a unit sphere, is
(i) 0
(iii) i1t (a+b+c)2
3
Ans. (ii)
(ii) 41t (a+b+c)
3
(iv) none of these
355
A Textbook of Engineering Mathematics Volume  I
Indicate True or False for the following Statements:
1. In case f is single valued, and integral is take round a closed curve, the
terminal points A and B coincide, and fs = fA
(True/False)
Ans.True
2. Green's theorem is a special case of Stoke's theorem
(True/False)
Ans. True
3. The Gauss divergence theorem is applicable for a region V if it is bounded by
two closed surfaces S1 and ~ one which lies within the other.
(True/False)
Ans.True
4. The value of f it x (i x r) ds is equal to zero where i is a constant vector.
(True/False)
Ans. False
5
5. For a closed surface S, the integral fit x V ~ d s vanishes identically
(True/False)
Ans. True
5
6. IfS be any closed surface, then f CurlF.dr = 0
(True/False)
Ans. True
5
7. If V represents the velocity of a fluid particle and C is a closed curve, then the
integral ~ V.dr is called the circulation of V round the curve C.
c
(True/False)
Ans. True
8. If the circulation of V round every closed curve in it region D Vanishes, then
V is said to be irrotational in D.
(True/False)
Ans. True
9. Any integral which is to be evaluated over a surface is called a surface integral.
(True/False)
Ans. True
10. Gauss divergence theorem is the relation between surface integral and line
integral.
(True/False)
Ans. False
356
Vector Integral Calculus
11. Gauss divergence theorem gives the relation between surface and volume
integrals.
(True/False)
Ans. True
12. Stokes theorem gives the relation between surface and volume integral.
(True/False)
Ans. False
13. Stoke's theorem gives the relation between line and surface integrals.
(True/False)
Ans.True
357
"This page is Intentionally Left Blank"